fix: suppress startup script logs in non-blocking SSH mode

In non-blocking mode (--wait=no), skip streaming startup script logs
while still showing the stage indicator and warnings for failures.

This eliminates log spam when using SSH via ProxyCommand or with
--wait=no, bringing behavior closer to standard SSH expectations.

Fixes #13580

.claude/docs/ARCHITECTURE.md

@@ -0,0 +1,126 @@
+# Coder Architecture
+
+This document provides an overview of Coder's architecture and core systems.
+
+## What is Coder?
+
+Coder is a platform for creating, managing, and using remote development environments (also known as Cloud Development Environments or CDEs). It leverages Terraform to define and provision these environments, which are referred to as "workspaces" within the project. The system is designed to be extensible, secure, and provide developers with a seamless remote development experience.
+
+## Core Architecture
+
+The heart of Coder is a control plane that orchestrates the creation and management of workspaces. This control plane interacts with separate Provisioner processes over gRPC to handle workspace builds. The Provisioners consume workspace definitions and use Terraform to create the actual infrastructure.
+
+The CLI package serves dual purposes - it can be used to launch the control plane itself and also provides client functionality for users to interact with an existing control plane instance. All user-facing frontend code is developed in TypeScript using React and lives in the `site/` directory.
+
+The database layer uses PostgreSQL with SQLC for generating type-safe database code. Database migrations are carefully managed to ensure both forward and backward compatibility through paired `.up.sql` and `.down.sql` files.
+
+## API Design
+
+Coder's API architecture combines REST and gRPC approaches. The REST API is defined in `coderd/coderd.go` and uses Chi for HTTP routing. This provides the primary interface for the frontend and external integrations.
+
+Internal communication with Provisioners occurs over gRPC, with service definitions maintained in `.proto` files. This separation allows for efficient binary communication with the components responsible for infrastructure management while providing a standard REST interface for human-facing applications.
+
+## Network Architecture
+
+Coder implements a secure networking layer based on Tailscale's Wireguard implementation. The `tailnet` package provides connectivity between workspace agents and clients through DERP (Designated Encrypted Relay for Packets) servers when direct connections aren't possible. This creates a secure overlay network allowing access to workspaces regardless of network topology, firewalls, or NAT configurations.
+
+### Tailnet and DERP System
+
+The networking system has three key components:
+
+1. **Tailnet**: An overlay network implemented in the `tailnet` package that provides secure, end-to-end encrypted connections between clients, the Coder server, and workspace agents.
+
+2. **DERP Servers**: These relay traffic when direct connections aren't possible. Coder provides several options:
+   - A built-in DERP server that runs on the Coder control plane
+   - Integration with Tailscale's global DERP infrastructure
+   - Support for custom DERP servers for lower latency or offline deployments
+
+3. **Direct Connections**: When possible, the system establishes peer-to-peer connections between clients and workspaces using STUN for NAT traversal. This requires both endpoints to send UDP traffic on ephemeral ports.
+
+### Workspace Proxies
+
+Workspace proxies (in the Enterprise edition) provide regional relay points for browser-based connections, reducing latency for geo-distributed teams. Key characteristics:
+
+- Deployed as independent servers that authenticate with the Coder control plane
+- Relay connections for SSH, workspace apps, port forwarding, and web terminals
+- Do not make direct database connections
+- Managed through the `coder wsproxy` commands
+- Implemented primarily in the `enterprise/wsproxy/` package
+
+## Agent System
+
+The workspace agent runs within each provisioned workspace and provides core functionality including:
+
+- SSH access to workspaces via the `agentssh` package
+- Port forwarding
+- Terminal connectivity via the `pty` package for pseudo-terminal support
+- Application serving
+- Healthcheck monitoring
+- Resource usage reporting
+
+Agents communicate with the control plane using the tailnet system and authenticate using secure tokens.
+
+## Workspace Applications
+
+Workspace applications (or "apps") provide browser-based access to services running within workspaces. The system supports:
+
+- HTTP(S) and WebSocket connections
+- Path-based or subdomain-based access URLs
+- Health checks to monitor application availability
+- Different sharing levels (owner-only, authenticated users, or public)
+- Custom icons and display settings
+
+The implementation is primarily in the `coderd/workspaceapps/` directory with components for URL generation, proxying connections, and managing application state.
+
+## Implementation Details
+
+The project structure separates frontend and backend concerns. React components and pages are organized in the `site/src/` directory, with Jest used for testing. The backend is primarily written in Go, with a strong emphasis on error handling patterns and test coverage.
+
+Database interactions are carefully managed through migrations in `coderd/database/migrations/` and queries in `coderd/database/queries/`. All new queries require proper database authorization (dbauthz) implementation to ensure that only users with appropriate permissions can access specific resources.
+
+## Authorization System
+
+The database authorization (dbauthz) system enforces fine-grained access control across all database operations. It uses role-based access control (RBAC) to validate user permissions before executing database operations. The `dbauthz` package wraps the database store and performs authorization checks before returning data. All database operations must pass through this layer to ensure security.
+
+## Testing Framework
+
+The codebase has a comprehensive testing approach with several key components:
+
+1. **Parallel Testing**: All tests must use `t.Parallel()` to run concurrently, which improves test suite performance and helps identify race conditions.
+
+2. **coderdtest Package**: This package in `coderd/coderdtest/` provides utilities for creating test instances of the Coder server, setting up test users and workspaces, and mocking external components.
+
+3. **Integration Tests**: Tests often span multiple components to verify system behavior, such as template creation, workspace provisioning, and agent connectivity.
+
+4. **Enterprise Testing**: Enterprise features have dedicated test utilities in the `coderdenttest` package.
+
+## Open Source and Enterprise Components
+
+The repository contains both open source and enterprise components:
+
+- Enterprise code lives primarily in the `enterprise/` directory
+- Enterprise features focus on governance, scalability (high availability), and advanced deployment options like workspace proxies
+- The boundary between open source and enterprise is managed through a licensing system
+- The same core codebase supports both editions, with enterprise features conditionally enabled
+
+## Development Philosophy
+
+Coder emphasizes clear error handling, with specific patterns required:
+
+- Concise error messages that avoid phrases like "failed to"
+- Wrapping errors with `%w` to maintain error chains
+- Using sentinel errors with the "err" prefix (e.g., `errNotFound`)
+
+All tests should run in parallel using `t.Parallel()` to ensure efficient testing and expose potential race conditions. The codebase is rigorously linted with golangci-lint to maintain consistent code quality.
+
+Git contributions follow a standard format with commit messages structured as `type: <message>`, where type is one of `feat`, `fix`, or `chore`.
+
+## Development Workflow
+
+Development can be initiated using `scripts/develop.sh` to start the application after making changes. Database schema updates should be performed through the migration system using `create_migration.sh <name>` to generate migration files, with each `.up.sql` migration paired with a corresponding `.down.sql` that properly reverts all changes.
+
+If the development database gets into a bad state, it can be completely reset by removing the PostgreSQL data directory with `rm -rf .coderv2/postgres`. This will destroy all data in the development database, requiring you to recreate any test users, templates, or workspaces after restarting the application.
+
+Code generation for the database layer uses `coderd/database/generate.sh`, and developers should refer to `sqlc.yaml` for the appropriate style and patterns to follow when creating new queries or tables.
+
+The focus should always be on maintaining security through proper database authorization, clean error handling, and comprehensive test coverage to ensure the platform remains robust and reliable.

.claude/docs/DATABASE.md

@@ -0,0 +1,218 @@
+# Database Development Patterns
+
+## Database Work Overview
+
+### Database Generation Process
+
+1. Modify SQL files in `coderd/database/queries/`
+2. Run `make gen`
+3. If errors about audit table, update `enterprise/audit/table.go`
+4. Run `make gen` again
+5. Run `make lint` to catch any remaining issues
+
+## Migration Guidelines
+
+### Creating Migration Files
+
+**Location**: `coderd/database/migrations/`
+**Format**: `{number}_{description}.{up|down}.sql`
+
+- Number must be unique and sequential
+- Always include both up and down migrations
+
+### Helper Scripts
+
+| Script                                                              | Purpose                                 |
+|---------------------------------------------------------------------|-----------------------------------------|
+| `./coderd/database/migrations/create_migration.sh "migration name"` | Creates new migration files             |
+| `./coderd/database/migrations/fix_migration_numbers.sh`             | Renumbers migrations to avoid conflicts |
+| `./coderd/database/migrations/create_fixture.sh "fixture name"`     | Creates test fixtures for migrations    |
+
+### Database Query Organization
+
+- **MUST DO**: Any changes to database - adding queries, modifying queries should be done in the `coderd/database/queries/*.sql` files
+- **MUST DO**: Queries are grouped in files relating to context - e.g. `prebuilds.sql`, `users.sql`, `oauth2.sql`
+- After making changes to any `coderd/database/queries/*.sql` files you must run `make gen` to generate respective ORM changes
+
+## Handling Nullable Fields
+
+Use `sql.NullString`, `sql.NullBool`, etc. for optional database fields:
+
+```go
+CodeChallenge: sql.NullString{
+    String: params.codeChallenge,
+    Valid:  params.codeChallenge != "",
+}
+```
+
+Set `.Valid = true` when providing values.
+
+## Audit Table Updates
+
+If adding fields to auditable types:
+
+1. Update `enterprise/audit/table.go`
+2. Add each new field with appropriate action:
+   - `ActionTrack`: Field should be tracked in audit logs
+   - `ActionIgnore`: Field should be ignored in audit logs
+   - `ActionSecret`: Field contains sensitive data
+3. Run `make gen` to verify no audit errors
+
+## Database Architecture
+
+### Core Components
+
+- **PostgreSQL 13+** recommended for production
+- **Migrations** managed with `migrate`
+- **Database authorization** through `dbauthz` package
+
+### Authorization Patterns
+
+```go
+// Public endpoints needing system access (OAuth2 registration)
+app, err := api.Database.GetOAuth2ProviderAppByClientID(dbauthz.AsSystemRestricted(ctx), clientID)
+
+// Authenticated endpoints with user context
+app, err := api.Database.GetOAuth2ProviderAppByClientID(ctx, clientID)
+
+// System operations in middleware
+roles, err := db.GetAuthorizationUserRoles(dbauthz.AsSystemRestricted(ctx), userID)
+```
+
+## Common Database Issues
+
+### Migration Issues
+
+1. **Migration conflicts**: Use `fix_migration_numbers.sh` to renumber
+2. **Missing down migration**: Always create both up and down files
+3. **Schema inconsistencies**: Verify against existing schema
+
+### Field Handling Issues
+
+1. **Nullable field errors**: Use `sql.Null*` types consistently
+2. **Missing audit entries**: Update `enterprise/audit/table.go`
+
+### Query Issues
+
+1. **Query organization**: Group related queries in appropriate files
+2. **Generated code errors**: Run `make gen` after query changes
+3. **Performance issues**: Add appropriate indexes in migrations
+
+## Database Testing
+
+### Test Database Setup
+
+```go
+func TestDatabaseFunction(t *testing.T) {
+    db := dbtestutil.NewDB(t)
+
+    // Test with real database
+    result, err := db.GetSomething(ctx, param)
+    require.NoError(t, err)
+    require.Equal(t, expected, result)
+}
+```
+
+## Best Practices
+
+### Schema Design
+
+1. **Use appropriate data types**: VARCHAR for strings, TIMESTAMP for times
+2. **Add constraints**: NOT NULL, UNIQUE, FOREIGN KEY as appropriate
+3. **Create indexes**: For frequently queried columns
+4. **Consider performance**: Normalize appropriately but avoid over-normalization
+
+### Query Writing
+
+1. **Use parameterized queries**: Prevent SQL injection
+2. **Handle errors appropriately**: Check for specific error types
+3. **Use transactions**: For related operations that must succeed together
+4. **Optimize queries**: Use EXPLAIN to understand query performance
+
+### Migration Writing
+
+1. **Make migrations reversible**: Always include down migration
+2. **Test migrations**: On copy of production data if possible
+3. **Keep migrations small**: One logical change per migration
+4. **Document complex changes**: Add comments explaining rationale
+
+## Advanced Patterns
+
+### Complex Queries
+
+```sql
+-- Example: Complex join with aggregation
+SELECT
+    u.id,
+    u.username,
+    COUNT(w.id) as workspace_count
+FROM users u
+LEFT JOIN workspaces w ON u.id = w.owner_id
+WHERE u.created_at > $1
+GROUP BY u.id, u.username
+ORDER BY workspace_count DESC;
+```
+
+### Conditional Queries
+
+```sql
+-- Example: Dynamic filtering
+SELECT * FROM oauth2_provider_apps
+WHERE
+    ($1::text IS NULL OR name ILIKE '%' || $1 || '%')
+    AND ($2::uuid IS NULL OR organization_id = $2)
+ORDER BY created_at DESC;
+```
+
+### Audit Patterns
+
+```go
+// Example: Auditable database operation
+func (q *sqlQuerier) UpdateUser(ctx context.Context, arg UpdateUserParams) (User, error) {
+    // Implementation here
+
+    // Audit the change
+    if auditor := audit.FromContext(ctx); auditor != nil {
+        auditor.Record(audit.UserUpdate{
+            UserID: arg.ID,
+            Old:    oldUser,
+            New:    newUser,
+        })
+    }
+
+    return newUser, nil
+}
+```
+
+## Debugging Database Issues
+
+### Common Debug Commands
+
+```bash
+# Check database connection
+make test-postgres
+
+# Run specific database tests
+go test ./coderd/database/... -run TestSpecificFunction
+
+# Check query generation
+make gen
+
+# Verify audit table
+make lint
+```
+
+### Debug Techniques
+
+1. **Enable query logging**: Set appropriate log levels
+2. **Use database tools**: pgAdmin, psql for direct inspection
+3. **Check constraints**: UNIQUE, FOREIGN KEY violations
+4. **Analyze performance**: Use EXPLAIN ANALYZE for slow queries
+
+### Troubleshooting Checklist
+
+- [ ] Migration files exist (both up and down)
+- [ ] `make gen` run after query changes
+- [ ] Audit table updated for new fields
+- [ ] Nullable fields use `sql.Null*` types
+- [ ] Authorization context appropriate for endpoint type

.claude/docs/OAUTH2.md

@@ -0,0 +1,157 @@
+# OAuth2 Development Guide
+
+## RFC Compliance Development
+
+### Implementing Standard Protocols
+
+When implementing standard protocols (OAuth2, OpenID Connect, etc.):
+
+1. **Fetch and Analyze Official RFCs**:
+   - Always read the actual RFC specifications before implementation
+   - Use WebFetch tool to get current RFC content for compliance verification
+   - Document RFC requirements in code comments
+
+2. **Default Values Matter**:
+   - Pay close attention to RFC-specified default values
+   - Example: RFC 7591 specifies `client_secret_basic` as default, not `client_secret_post`
+   - Ensure consistency between database migrations and application code
+
+3. **Security Requirements**:
+   - Follow RFC security considerations precisely
+   - Example: RFC 7592 prohibits returning registration access tokens in GET responses
+   - Implement proper error responses per protocol specifications
+
+4. **Validation Compliance**:
+   - Implement comprehensive validation per RFC requirements
+   - Support protocol-specific features (e.g., custom schemes for native OAuth2 apps)
+   - Test edge cases defined in specifications
+
+## OAuth2 Provider Implementation
+
+### OAuth2 Spec Compliance
+
+1. **Follow RFC 6749 for token responses**
+   - Use `expires_in` (seconds) not `expiry` (timestamp) in token responses
+   - Return proper OAuth2 error format: `{"error": "code", "error_description": "details"}`
+
+2. **Error Response Format**
+   - Create OAuth2-compliant error responses for token endpoint
+   - Use standard error codes: `invalid_client`, `invalid_grant`, `invalid_request`
+   - Avoid generic error responses for OAuth2 endpoints
+
+### PKCE Implementation
+
+- Support both with and without PKCE for backward compatibility
+- Use S256 method for code challenge
+- Properly validate code_verifier against stored code_challenge
+
+### UI Authorization Flow
+
+- Use POST requests for consent, not GET with links
+- Avoid dependency on referer headers for security decisions
+- Support proper state parameter validation
+
+### RFC 8707 Resource Indicators
+
+- Store resource parameters in database for server-side validation (opaque tokens)
+- Validate resource consistency between authorization and token requests
+- Support audience validation in refresh token flows
+- Resource parameter is optional but must be consistent when provided
+
+## OAuth2 Error Handling Pattern
+
+```go
+// Define specific OAuth2 errors
+var (
+    errInvalidPKCE = xerrors.New("invalid code_verifier")
+)
+
+// Use OAuth2-compliant error responses
+type OAuth2Error struct {
+    Error            string `json:"error"`
+    ErrorDescription string `json:"error_description,omitempty"`
+}
+
+// Return proper OAuth2 errors
+if errors.Is(err, errInvalidPKCE) {
+    writeOAuth2Error(ctx, rw, http.StatusBadRequest, "invalid_grant", "The PKCE code verifier is invalid")
+    return
+}
+```
+
+## Testing OAuth2 Features
+
+### Test Scripts
+
+Located in `./scripts/oauth2/`:
+
+- `test-mcp-oauth2.sh` - Full automated test suite
+- `setup-test-app.sh` - Create test OAuth2 app
+- `cleanup-test-app.sh` - Remove test app
+- `generate-pkce.sh` - Generate PKCE parameters
+- `test-manual-flow.sh` - Manual browser testing
+
+Always run the full test suite after OAuth2 changes:
+
+```bash
+./scripts/oauth2/test-mcp-oauth2.sh
+```
+
+### RFC Protocol Testing
+
+1. **Compliance Test Coverage**:
+   - Test all RFC-defined error codes and responses
+   - Validate proper HTTP status codes for different scenarios
+   - Test protocol-specific edge cases (URI formats, token formats, etc.)
+
+2. **Security Boundary Testing**:
+   - Test client isolation and privilege separation
+   - Verify information disclosure protections
+   - Test token security and proper invalidation
+
+## Common OAuth2 Issues
+
+1. **OAuth2 endpoints returning wrong error format** - Ensure OAuth2 endpoints return RFC 6749 compliant errors
+2. **Resource indicator validation failing** - Ensure database stores and retrieves resource parameters correctly
+3. **PKCE tests failing** - Verify both authorization code storage and token exchange handle PKCE fields
+4. **RFC compliance failures** - Verify against actual RFC specifications, not assumptions
+5. **Authorization context errors in public endpoints** - Use `dbauthz.AsSystemRestricted(ctx)` pattern
+6. **Default value mismatches** - Ensure database migrations match application code defaults
+7. **Bearer token authentication issues** - Check token extraction precedence and format validation
+8. **URI validation failures** - Support both standard schemes and custom schemes per protocol requirements
+
+## Authorization Context Patterns
+
+```go
+// Public endpoints needing system access (OAuth2 registration)
+app, err := api.Database.GetOAuth2ProviderAppByClientID(dbauthz.AsSystemRestricted(ctx), clientID)
+
+// Authenticated endpoints with user context
+app, err := api.Database.GetOAuth2ProviderAppByClientID(ctx, clientID)
+
+// System operations in middleware
+roles, err := db.GetAuthorizationUserRoles(dbauthz.AsSystemRestricted(ctx), userID)
+```
+
+## OAuth2/Authentication Work Patterns
+
+- Types go in `codersdk/oauth2.go` or similar
+- Handlers go in `coderd/oauth2.go` or `coderd/identityprovider/`
+- Database fields need migration + audit table updates
+- Always support backward compatibility
+
+## Protocol Implementation Checklist
+
+Before completing OAuth2 or authentication feature work:
+
+- [ ] Verify RFC compliance by reading actual specifications
+- [ ] Implement proper error response formats per protocol
+- [ ] Add comprehensive validation for all protocol fields
+- [ ] Test security boundaries and token handling
+- [ ] Update RBAC permissions for new resources
+- [ ] Add audit logging support if applicable
+- [ ] Create database migrations with proper defaults
+- [ ] Add comprehensive test coverage including edge cases
+- [ ] Verify linting compliance
+- [ ] Test both positive and negative scenarios
+- [ ] Document protocol-specific patterns and requirements

.claude/docs/TESTING.md

@@ -0,0 +1,212 @@
+# Testing Patterns and Best Practices
+
+## Testing Best Practices
+
+### Avoiding Race Conditions
+
+1. **Unique Test Identifiers**:
+   - Never use hardcoded names in concurrent tests
+   - Use `time.Now().UnixNano()` or similar for unique identifiers
+   - Example: `fmt.Sprintf("test-client-%s-%d", t.Name(), time.Now().UnixNano())`
+
+2. **Database Constraint Awareness**:
+   - Understand unique constraints that can cause test conflicts
+   - Generate unique values for all constrained fields
+   - Test name isolation prevents cross-test interference
+
+### Testing Patterns
+
+- Use table-driven tests for comprehensive coverage
+- Mock external dependencies
+- Test both positive and negative cases
+- Use `testutil.WaitLong` for timeouts in tests
+
+### Test Package Naming
+
+- **Test packages**: Use `package_test` naming (e.g., `identityprovider_test`) for black-box testing
+
+## RFC Protocol Testing
+
+### Compliance Test Coverage
+
+1. **Test all RFC-defined error codes and responses**
+2. **Validate proper HTTP status codes for different scenarios**
+3. **Test protocol-specific edge cases** (URI formats, token formats, etc.)
+
+### Security Boundary Testing
+
+1. **Test client isolation and privilege separation**
+2. **Verify information disclosure protections**
+3. **Test token security and proper invalidation**
+
+## Test Organization
+
+### Test File Structure
+
+```
+coderd/
+├── oauth2.go                    # Implementation
+├── oauth2_test.go              # Main tests
+├── oauth2_test_helpers.go      # Test utilities
+└── oauth2_validation.go        # Validation logic
+```
+
+### Test Categories
+
+1. **Unit Tests**: Test individual functions in isolation
+2. **Integration Tests**: Test API endpoints with database
+3. **End-to-End Tests**: Full workflow testing
+4. **Race Tests**: Concurrent access testing
+
+## Test Commands
+
+### Running Tests
+
+| Command | Purpose |
+|---------|---------|
+| `make test` | Run all Go tests |
+| `make test RUN=TestFunctionName` | Run specific test |
+| `go test -v ./path/to/package -run TestFunctionName` | Run test with verbose output |
+| `make test-postgres` | Run tests with Postgres database |
+| `make test-race` | Run tests with Go race detector |
+| `make test-e2e` | Run end-to-end tests |
+
+### Frontend Testing
+
+| Command | Purpose |
+|---------|---------|
+| `pnpm test` | Run frontend tests |
+| `pnpm check` | Run code checks |
+
+## Common Testing Issues
+
+### Database-Related
+
+1. **SQL type errors** - Use `sql.Null*` types for nullable fields
+2. **Race conditions in tests** - Use unique identifiers instead of hardcoded names
+
+### OAuth2 Testing
+
+1. **PKCE tests failing** - Verify both authorization code storage and token exchange handle PKCE fields
+2. **Resource indicator validation failing** - Ensure database stores and retrieves resource parameters correctly
+
+### General Issues
+
+1. **Missing newlines** - Ensure files end with newline character
+2. **Package naming errors** - Use `package_test` naming for test files
+3. **Log message formatting errors** - Use lowercase, descriptive messages without special characters
+
+## Systematic Testing Approach
+
+### Multi-Issue Problem Solving
+
+When facing multiple failing tests or complex integration issues:
+
+1. **Identify Root Causes**:
+   - Run failing tests individually to isolate issues
+   - Use LSP tools to trace through call chains
+   - Check both compilation and runtime errors
+
+2. **Fix in Logical Order**:
+   - Address compilation issues first (imports, syntax)
+   - Fix authorization and RBAC issues next
+   - Resolve business logic and validation issues
+   - Handle edge cases and race conditions last
+
+3. **Verification Strategy**:
+   - Test each fix individually before moving to next issue
+   - Use `make lint` and `make gen` after database changes
+   - Verify RFC compliance with actual specifications
+   - Run comprehensive test suites before considering complete
+
+## Test Data Management
+
+### Unique Test Data
+
+```go
+// Good: Unique identifiers prevent conflicts
+clientName := fmt.Sprintf("test-client-%s-%d", t.Name(), time.Now().UnixNano())
+
+// Bad: Hardcoded names cause race conditions
+clientName := "test-client"
+```
+
+### Test Cleanup
+
+```go
+func TestSomething(t *testing.T) {
+    // Setup
+    client := coderdtest.New(t, nil)
+
+    // Test code here
+
+    // Cleanup happens automatically via t.Cleanup() in coderdtest
+}
+```
+
+## Test Utilities
+
+### Common Test Patterns
+
+```go
+// Table-driven tests
+tests := []struct {
+    name     string
+    input    InputType
+    expected OutputType
+    wantErr  bool
+}{
+    {
+        name:     "valid input",
+        input:    validInput,
+        expected: expectedOutput,
+        wantErr:  false,
+    },
+    // ... more test cases
+}
+
+for _, tt := range tests {
+    t.Run(tt.name, func(t *testing.T) {
+        result, err := functionUnderTest(tt.input)
+        if tt.wantErr {
+            require.Error(t, err)
+            return
+        }
+        require.NoError(t, err)
+        require.Equal(t, tt.expected, result)
+    })
+}
+```
+
+### Test Assertions
+
+```go
+// Use testify/require for assertions
+require.NoError(t, err)
+require.Equal(t, expected, actual)
+require.NotNil(t, result)
+require.True(t, condition)
+```
+
+## Performance Testing
+
+### Load Testing
+
+- Use `scaletest/` directory for load testing scenarios
+- Run `./scaletest/scaletest.sh` for performance testing
+
+### Benchmarking
+
+```go
+func BenchmarkFunction(b *testing.B) {
+    for i := 0; i < b.N; i++ {
+        // Function call to benchmark
+        _ = functionUnderTest(input)
+    }
+}
+```
+
+Run benchmarks with:
+```bash
+go test -bench=. -benchmem ./package/path
+```

.claude/docs/TROUBLESHOOTING.md

@@ -0,0 +1,239 @@
+# Troubleshooting Guide
+
+## Common Issues
+
+### Database Issues
+
+1. **"Audit table entry missing action"**
+   - **Solution**: Update `enterprise/audit/table.go`
+   - Add each new field with appropriate action (ActionTrack, ActionIgnore, ActionSecret)
+   - Run `make gen` to verify no audit errors
+
+2. **SQL type errors**
+   - **Solution**: Use `sql.Null*` types for nullable fields
+   - Set `.Valid = true` when providing values
+   - Example:
+
+     ```go
+     CodeChallenge: sql.NullString{
+         String: params.codeChallenge,
+         Valid:  params.codeChallenge != "",
+     }
+     ```
+
+### Testing Issues
+
+3. **"package should be X_test"**
+   - **Solution**: Use `package_test` naming for test files
+   - Example: `identityprovider_test` for black-box testing
+
+4. **Race conditions in tests**
+   - **Solution**: Use unique identifiers instead of hardcoded names
+   - Example: `fmt.Sprintf("test-client-%s-%d", t.Name(), time.Now().UnixNano())`
+   - Never use hardcoded names in concurrent tests
+
+5. **Missing newlines**
+   - **Solution**: Ensure files end with newline character
+   - Most editors can be configured to add this automatically
+
+### OAuth2 Issues
+
+6. **OAuth2 endpoints returning wrong error format**
+   - **Solution**: Ensure OAuth2 endpoints return RFC 6749 compliant errors
+   - Use standard error codes: `invalid_client`, `invalid_grant`, `invalid_request`
+   - Format: `{"error": "code", "error_description": "details"}`
+
+7. **Resource indicator validation failing**
+   - **Solution**: Ensure database stores and retrieves resource parameters correctly
+   - Check both authorization code storage and token exchange handling
+
+8. **PKCE tests failing**
+    - **Solution**: Verify both authorization code storage and token exchange handle PKCE fields
+    - Check `CodeChallenge` and `CodeChallengeMethod` field handling
+
+### RFC Compliance Issues
+
+9. **RFC compliance failures**
+    - **Solution**: Verify against actual RFC specifications, not assumptions
+    - Use WebFetch tool to get current RFC content for compliance verification
+    - Read the actual RFC specifications before implementation
+
+10. **Default value mismatches**
+    - **Solution**: Ensure database migrations match application code defaults
+    - Example: RFC 7591 specifies `client_secret_basic` as default, not `client_secret_post`
+
+### Authorization Issues
+
+11. **Authorization context errors in public endpoints**
+    - **Solution**: Use `dbauthz.AsSystemRestricted(ctx)` pattern
+    - Example:
+
+      ```go
+      // Public endpoints needing system access
+      app, err := api.Database.GetOAuth2ProviderAppByClientID(dbauthz.AsSystemRestricted(ctx), clientID)
+      ```
+
+### Authentication Issues
+
+12. **Bearer token authentication issues**
+    - **Solution**: Check token extraction precedence and format validation
+    - Ensure proper RFC 6750 Bearer Token Support implementation
+
+13. **URI validation failures**
+    - **Solution**: Support both standard schemes and custom schemes per protocol requirements
+    - Native OAuth2 apps may use custom schemes
+
+### General Development Issues
+
+14. **Log message formatting errors**
+    - **Solution**: Use lowercase, descriptive messages without special characters
+    - Follow Go logging conventions
+
+## Systematic Debugging Approach
+
+YOU MUST ALWAYS find the root cause of any issue you are debugging
+YOU MUST NEVER fix a symptom or add a workaround instead of finding a root cause, even if it is faster.
+
+### Multi-Issue Problem Solving
+
+When facing multiple failing tests or complex integration issues:
+
+1. **Identify Root Causes**:
+   - Run failing tests individually to isolate issues
+   - Use LSP tools to trace through call chains
+   - Read Error Messages Carefully: Check both compilation and runtime errors
+   - Reproduce Consistently: Ensure you can reliably reproduce the issue before investigating
+   - Check Recent Changes: What changed that could have caused this? Git diff, recent commits, etc.
+   - When You Don't Know: Say "I don't understand X" rather than pretending to know
+
+2. **Fix in Logical Order**:
+   - Address compilation issues first (imports, syntax)
+   - Fix authorization and RBAC issues next
+   - Resolve business logic and validation issues
+   - Handle edge cases and race conditions last
+   - IF your first fix doesn't work, STOP and re-analyze rather than adding more fixes
+
+3. **Verification Strategy**:
+   - Always Test each fix individually before moving to next issue
+   - Verify Before Continuing: Did your test work? If not, form new hypothesis - don't add more fixes
+   - Use `make lint` and `make gen` after database changes
+   - Verify RFC compliance with actual specifications
+   - Run comprehensive test suites before considering complete
+
+## Debug Commands
+
+### Useful Debug Commands
+
+| Command                                      | Purpose                               |
+|----------------------------------------------|---------------------------------------|
+| `make lint`                                  | Run all linters                       |
+| `make gen`                                   | Generate mocks, database queries      |
+| `go test -v ./path/to/package -run TestName` | Run specific test with verbose output |
+| `go test -race ./...`                        | Run tests with race detector          |
+
+### LSP Debugging
+
+#### Go LSP (Backend)
+
+| Command                                            | Purpose                      |
+|----------------------------------------------------|------------------------------|
+| `mcp__go-language-server__definition symbolName`   | Find function definition     |
+| `mcp__go-language-server__references symbolName`   | Find all references          |
+| `mcp__go-language-server__diagnostics filePath`    | Check for compilation errors |
+| `mcp__go-language-server__hover filePath line col` | Get type information         |
+
+#### TypeScript LSP (Frontend)
+
+| Command                                                                    | Purpose                            |
+|----------------------------------------------------------------------------|------------------------------------|
+| `mcp__typescript-language-server__definition symbolName`                   | Find component/function definition |
+| `mcp__typescript-language-server__references symbolName`                   | Find all component/type usages     |
+| `mcp__typescript-language-server__diagnostics filePath`                    | Check for TypeScript errors        |
+| `mcp__typescript-language-server__hover filePath line col`                 | Get type information               |
+| `mcp__typescript-language-server__rename_symbol filePath line col newName` | Rename across codebase             |
+
+## Common Error Messages
+
+### Database Errors
+
+**Error**: `pq: relation "oauth2_provider_app_codes" does not exist`
+
+- **Cause**: Missing database migration
+- **Solution**: Run database migrations, check migration files
+
+**Error**: `audit table entry missing action for field X`
+
+- **Cause**: New field added without audit table update
+- **Solution**: Update `enterprise/audit/table.go`
+
+### Go Compilation Errors
+
+**Error**: `package should be identityprovider_test`
+
+- **Cause**: Test package naming convention violation
+- **Solution**: Use `package_test` naming for black-box tests
+
+**Error**: `cannot use X (type Y) as type Z`
+
+- **Cause**: Type mismatch, often with nullable fields
+- **Solution**: Use appropriate `sql.Null*` types
+
+### OAuth2 Errors
+
+**Error**: `invalid_client` but client exists
+
+- **Cause**: Authorization context issue
+- **Solution**: Use `dbauthz.AsSystemRestricted(ctx)` for public endpoints
+
+**Error**: PKCE validation failing
+
+- **Cause**: Missing PKCE fields in database operations
+- **Solution**: Ensure `CodeChallenge` and `CodeChallengeMethod` are handled
+
+## Prevention Strategies
+
+### Before Making Changes
+
+1. **Read the relevant documentation**
+2. **Check if similar patterns exist in codebase**
+3. **Understand the authorization context requirements**
+4. **Plan database changes carefully**
+
+### During Development
+
+1. **Run tests frequently**: `make test`
+2. **Use LSP tools for navigation**: Avoid manual searching
+3. **Follow RFC specifications precisely**
+4. **Update audit tables when adding database fields**
+
+### Before Committing
+
+1. **Run full test suite**: `make test`
+2. **Check linting**: `make lint`
+3. **Test with race detector**: `make test-race`
+
+## Getting Help
+
+### Internal Resources
+
+- Check existing similar implementations in codebase
+- Use LSP tools to understand code relationships
+  - For Go code: Use `mcp__go-language-server__*` commands
+  - For TypeScript/React code: Use `mcp__typescript-language-server__*` commands
+- Read related test files for expected behavior
+
+### External Resources
+
+- Official RFC specifications for protocol compliance
+- Go documentation for language features
+- PostgreSQL documentation for database issues
+
+### Debug Information Collection
+
+When reporting issues, include:
+
+1. **Exact error message**
+2. **Steps to reproduce**
+3. **Relevant code snippets**
+4. **Test output (if applicable)**
+5. **Environment information** (OS, Go version, etc.)

.claude/docs/WORKFLOWS.md

@@ -0,0 +1,227 @@
+# Development Workflows and Guidelines
+
+## Quick Start Checklist for New Features
+
+### Before Starting
+
+- [ ] Run `git pull` to ensure you're on latest code
+- [ ] Check if feature touches database - you'll need migrations
+- [ ] Check if feature touches audit logs - update `enterprise/audit/table.go`
+
+## Development Server
+
+### Starting Development Mode
+
+- **Use `./scripts/develop.sh` to start Coder in development mode**
+- This automatically builds and runs with `--dev` flag and proper access URL
+- **⚠️ Do NOT manually run `make build && ./coder server --dev` - use the script instead**
+
+### Development Workflow
+
+1. **Always start with the development script**: `./scripts/develop.sh`
+2. **Make changes** to your code
+3. **The script will automatically rebuild** and restart as needed
+4. **Access the development server** at the URL provided by the script
+
+## Code Style Guidelines
+
+### Go Style
+
+- Follow [Effective Go](https://go.dev/doc/effective_go) and [Go's Code Review Comments](https://github.com/golang/go/wiki/CodeReviewComments)
+- Create packages when used during implementation
+- Validate abstractions against implementations
+- **Test packages**: Use `package_test` naming (e.g., `identityprovider_test`) for black-box testing
+
+### Error Handling
+
+- Use descriptive error messages
+- Wrap errors with context
+- Propagate errors appropriately
+- Use proper error types
+- Pattern: `xerrors.Errorf("failed to X: %w", err)`
+
+## Naming Conventions
+
+- Names MUST tell what code does, not how it's implemented or its history
+- Follow Go and TypeScript naming conventions
+- When changing code, never document the old behavior or the behavior change
+- NEVER use implementation details in names (e.g., "ZodValidator", "MCPWrapper", "JSONParser")
+- NEVER use temporal/historical context in names (e.g., "LegacyHandler", "UnifiedTool", "ImprovedInterface", "EnhancedParser")
+- NEVER use pattern names unless they add clarity (e.g., prefer "Tool" over "ToolFactory")
+- Abbreviate only when obvious
+
+### Comments
+
+- Document exported functions, types, and non-obvious logic
+- Follow JSDoc format for TypeScript
+- Use godoc format for Go code
+
+## Database Migration Workflows
+
+### Migration Guidelines
+
+1. **Create migration files**:
+   - Location: `coderd/database/migrations/`
+   - Format: `{number}_{description}.{up|down}.sql`
+   - Number must be unique and sequential
+   - Always include both up and down migrations
+
+2. **Use helper scripts**:
+   - `./coderd/database/migrations/create_migration.sh "migration name"` - Creates new migration files
+   - `./coderd/database/migrations/fix_migration_numbers.sh` - Renumbers migrations to avoid conflicts
+   - `./coderd/database/migrations/create_fixture.sh "fixture name"` - Creates test fixtures for migrations
+
+3. **Update database queries**:
+   - **MUST DO**: Any changes to database - adding queries, modifying queries should be done in the `coderd/database/queries/*.sql` files
+   - **MUST DO**: Queries are grouped in files relating to context - e.g. `prebuilds.sql`, `users.sql`, `oauth2.sql`
+   - After making changes to any `coderd/database/queries/*.sql` files you must run `make gen` to generate respective ORM changes
+
+4. **Handle nullable fields**:
+   - Use `sql.NullString`, `sql.NullBool`, etc. for optional database fields
+   - Set `.Valid = true` when providing values
+
+5. **Audit table updates**:
+   - If adding fields to auditable types, update `enterprise/audit/table.go`
+   - Add each new field with appropriate action (ActionTrack, ActionIgnore, ActionSecret)
+   - Run `make gen` to verify no audit errors
+
+### Database Generation Process
+
+1. Modify SQL files in `coderd/database/queries/`
+2. Run `make gen`
+3. If errors about audit table, update `enterprise/audit/table.go`
+4. Run `make gen` again
+5. Run `make lint` to catch any remaining issues
+
+## API Development Workflow
+
+### Adding New API Endpoints
+
+1. **Define types** in `codersdk/` package
+2. **Add handler** in appropriate `coderd/` file
+3. **Register route** in `coderd/coderd.go`
+4. **Add tests** in `coderd/*_test.go` files
+5. **Update OpenAPI** by running `make gen`
+
+## Testing Workflows
+
+### Test Execution
+
+- Run full test suite: `make test`
+- Run specific test: `make test RUN=TestFunctionName`
+- Run with Postgres: `make test-postgres`
+- Run with race detector: `make test-race`
+- Run end-to-end tests: `make test-e2e`
+
+### Test Development
+
+- Use table-driven tests for comprehensive coverage
+- Mock external dependencies
+- Test both positive and negative cases
+- Use `testutil.WaitLong` for timeouts in tests
+- Always use `t.Parallel()` in tests
+
+## Commit Style
+
+- Follow [Conventional Commits 1.0.0](https://www.conventionalcommits.org/en/v1.0.0/)
+- Format: `type(scope): message`
+- Types: `feat`, `fix`, `docs`, `style`, `refactor`, `test`, `chore`
+- Keep message titles concise (~70 characters)
+- Use imperative, present tense in commit titles
+
+## Code Navigation and Investigation
+
+### Using LSP Tools (STRONGLY RECOMMENDED)
+
+**IMPORTANT**: Always use LSP tools for code navigation and understanding. These tools provide accurate, real-time analysis of the codebase and should be your first choice for code investigation.
+
+#### Go LSP Tools (for backend code)
+
+1. **Find function definitions** (USE THIS FREQUENTLY):
+   - `mcp__go-language-server__definition symbolName`
+   - Example: `mcp__go-language-server__definition getOAuth2ProviderAppAuthorize`
+   - Quickly jump to function implementations across packages
+
+2. **Find symbol references** (ESSENTIAL FOR UNDERSTANDING IMPACT):
+   - `mcp__go-language-server__references symbolName`
+   - Locate all usages of functions, types, or variables
+   - Critical for refactoring and understanding data flow
+
+3. **Get symbol information**:
+   - `mcp__go-language-server__hover filePath line column`
+   - Get type information and documentation at specific positions
+
+#### TypeScript LSP Tools (for frontend code in site/)
+
+1. **Find component/function definitions** (USE THIS FREQUENTLY):
+   - `mcp__typescript-language-server__definition symbolName`
+   - Example: `mcp__typescript-language-server__definition LoginPage`
+   - Quickly navigate to React components, hooks, and utility functions
+
+2. **Find symbol references** (ESSENTIAL FOR UNDERSTANDING IMPACT):
+   - `mcp__typescript-language-server__references symbolName`
+   - Locate all usages of components, types, or functions
+   - Critical for refactoring React components and understanding prop usage
+
+3. **Get type information**:
+   - `mcp__typescript-language-server__hover filePath line column`
+   - Get TypeScript type information and JSDoc documentation
+
+4. **Rename symbols safely**:
+   - `mcp__typescript-language-server__rename_symbol filePath line column newName`
+   - Rename components, props, or functions across the entire codebase
+
+5. **Check for TypeScript errors**:
+   - `mcp__typescript-language-server__diagnostics filePath`
+   - Get compilation errors and warnings for a specific file
+
+### Investigation Strategy (LSP-First Approach)
+
+#### Backend Investigation (Go)
+
+1. **Start with route registration** in `coderd/coderd.go` to understand API endpoints
+2. **Use Go LSP `definition` lookup** to trace from route handlers to actual implementations
+3. **Use Go LSP `references`** to understand how functions are called throughout the codebase
+4. **Follow the middleware chain** using LSP tools to understand request processing flow
+5. **Check test files** for expected behavior and error patterns
+
+#### Frontend Investigation (TypeScript/React)
+
+1. **Start with route definitions** in `site/src/App.tsx` or router configuration
+2. **Use TypeScript LSP `definition`** to navigate to React components and hooks
+3. **Use TypeScript LSP `references`** to find all component usages and prop drilling
+4. **Follow the component hierarchy** using LSP tools to understand data flow
+5. **Check for TypeScript errors** with `diagnostics` before making changes
+6. **Examine test files** (`.test.tsx`) for component behavior and expected props
+
+## Troubleshooting Development Issues
+
+### Common Issues
+
+1. **Development server won't start** - Use `./scripts/develop.sh` instead of manual commands
+2. **Database migration errors** - Check migration file format and use helper scripts
+3. **Audit table errors** - Update `enterprise/audit/table.go` with new fields
+4. **OAuth2 compliance issues** - Ensure RFC-compliant error responses
+
+### Debug Commands
+
+- Check linting: `make lint`
+- Generate code: `make gen`
+- Clean build: `make clean`
+
+## Development Environment Setup
+
+### Prerequisites
+
+- Go (version specified in go.mod)
+- Node.js and pnpm for frontend development
+- PostgreSQL for database testing
+- Docker for containerized testing
+
+### First Time Setup
+
+1. Clone the repository
+2. Run `./scripts/develop.sh` to start development server
+3. Access the development URL provided
+4. Create admin user as prompted
+5. Begin development

.claude/scripts/format.sh

@@ -0,0 +1,133 @@
+#!/bin/bash
+
+# Claude Code hook script for file formatting
+# This script integrates with the centralized Makefile formatting targets
+# and supports the Claude Code hooks system for automatic file formatting.
+
+set -euo pipefail
+
+# A variable to memoize the command for canonicalizing paths.
+_CANONICALIZE_CMD=""
+
+# canonicalize_path resolves a path to its absolute, canonical form.
+# It tries 'realpath' and 'readlink -f' in order.
+# The chosen command is memoized to avoid repeated checks.
+# If none of these are available, it returns an empty string.
+canonicalize_path() {
+	local path_to_resolve="$1"
+
+	# If we haven't determined a command yet, find one.
+	if [[ -z "$_CANONICALIZE_CMD" ]]; then
+		if command -v realpath >/dev/null 2>&1; then
+			_CANONICALIZE_CMD="realpath"
+		elif command -v readlink >/dev/null 2>&1 && readlink -f . >/dev/null 2>&1; then
+			_CANONICALIZE_CMD="readlink"
+		else
+			# No command found, so we can't resolve.
+			# We set a "none" value to prevent re-checking.
+			_CANONICALIZE_CMD="none"
+		fi
+	fi
+
+	# Now, execute the command.
+	case "$_CANONICALIZE_CMD" in
+	realpath)
+		realpath "$path_to_resolve" 2>/dev/null
+		;;
+	readlink)
+		readlink -f "$path_to_resolve" 2>/dev/null
+		;;
+	*)
+		# This handles the "none" case or any unexpected error.
+		echo ""
+		;;
+	esac
+}
+
+# Read JSON input from stdin
+input=$(cat)
+
+# Extract the file path from the JSON input
+# Expected format: {"tool_input": {"file_path": "/absolute/path/to/file"}} or {"tool_response": {"filePath": "/absolute/path/to/file"}}
+file_path=$(echo "$input" | jq -r '.tool_input.file_path // .tool_response.filePath // empty')
+
+# Secure path canonicalization to prevent path traversal attacks
+# Resolve repo root to an absolute, canonical path.
+repo_root_raw="$(cd "$(dirname "$0")/../.." && pwd)"
+repo_root="$(canonicalize_path "$repo_root_raw")"
+if [[ -z "$repo_root" ]]; then
+	# Fallback if canonicalization fails
+	repo_root="$repo_root_raw"
+fi
+
+# Resolve the input path to an absolute path
+if [[ "$file_path" = /* ]]; then
+	# Already absolute
+	abs_file_path="$file_path"
+else
+	# Make relative paths absolute from repo root
+	abs_file_path="$repo_root/$file_path"
+fi
+
+# Canonicalize the path (resolve symlinks and ".." segments)
+canonical_file_path="$(canonicalize_path "$abs_file_path")"
+
+# Check if canonicalization failed or if the resolved path is outside the repo
+if [[ -z "$canonical_file_path" ]] || { [[ "$canonical_file_path" != "$repo_root" ]] && [[ "$canonical_file_path" != "$repo_root"/* ]]; }; then
+	echo "Error: File path is outside repository or invalid: $file_path" >&2
+	exit 1
+fi
+
+# Handle the case where the file path is the repository root itself.
+if [[ "$canonical_file_path" == "$repo_root" ]]; then
+	echo "Warning: Formatting the repository root is not a supported operation. Skipping." >&2
+	exit 0
+fi
+
+# Convert back to relative path from repo root for consistency
+file_path="${canonical_file_path#"$repo_root"/}"
+
+if [[ -z "$file_path" ]]; then
+	echo "Error: No file path provided in input" >&2
+	exit 1
+fi
+
+# Check if file exists
+if [[ ! -f "$file_path" ]]; then
+	echo "Error: File does not exist: $file_path" >&2
+	exit 1
+fi
+
+# Get the file extension to determine the appropriate formatter
+file_ext="${file_path##*.}"
+
+# Change to the project root directory (where the Makefile is located)
+cd "$(dirname "$0")/../.."
+
+# Call the appropriate Makefile target based on file extension
+case "$file_ext" in
+go)
+	make fmt/go FILE="$file_path"
+	echo "✓ Formatted Go file: $file_path"
+	;;
+js | jsx | ts | tsx)
+	make fmt/ts FILE="$file_path"
+	echo "✓ Formatted TypeScript/JavaScript file: $file_path"
+	;;
+tf | tfvars)
+	make fmt/terraform FILE="$file_path"
+	echo "✓ Formatted Terraform file: $file_path"
+	;;
+sh)
+	make fmt/shfmt FILE="$file_path"
+	echo "✓ Formatted shell script: $file_path"
+	;;
+md)
+	make fmt/markdown FILE="$file_path"
+	echo "✓ Formatted Markdown file: $file_path"
+	;;
+*)
+	echo "No formatter available for file extension: $file_ext"
+	exit 0
+	;;
+esac

.claude/settings.json

@@ -0,0 +1,15 @@
+{
+	"hooks": {
+		"PostToolUse": [
+			{
+				"matcher": "Edit|Write|MultiEdit",
+				"hooks": [
+					{
+						"type": "command",
+						"command": ".claude/scripts/format.sh"
+					}
+				]
+			}
+		]
+	}
+}

.cursorrules

@@ -1,124 +0,0 @@
-# Cursor Rules
-
-This project is called "Coder" - an application for managing remote development environments.
-
-Coder provides a platform for creating, managing, and using remote development environments (also known as Cloud Development Environments or CDEs). It leverages Terraform to define and provision these environments, which are referred to as "workspaces" within the project. The system is designed to be extensible, secure, and provide developers with a seamless remote development experience.
-
-## Core Architecture
-
-The heart of Coder is a control plane that orchestrates the creation and management of workspaces. This control plane interacts with separate Provisioner processes over gRPC to handle workspace builds. The Provisioners consume workspace definitions and use Terraform to create the actual infrastructure.
-
-The CLI package serves dual purposes - it can be used to launch the control plane itself and also provides client functionality for users to interact with an existing control plane instance. All user-facing frontend code is developed in TypeScript using React and lives in the `site/` directory.
-
-The database layer uses PostgreSQL with SQLC for generating type-safe database code. Database migrations are carefully managed to ensure both forward and backward compatibility through paired `.up.sql` and `.down.sql` files.
-
-## API Design
-
-Coder's API architecture combines REST and gRPC approaches. The REST API is defined in `coderd/coderd.go` and uses Chi for HTTP routing. This provides the primary interface for the frontend and external integrations.
-
-Internal communication with Provisioners occurs over gRPC, with service definitions maintained in `.proto` files. This separation allows for efficient binary communication with the components responsible for infrastructure management while providing a standard REST interface for human-facing applications.
-
-## Network Architecture
-
-Coder implements a secure networking layer based on Tailscale's Wireguard implementation. The `tailnet` package provides connectivity between workspace agents and clients through DERP (Designated Encrypted Relay for Packets) servers when direct connections aren't possible. This creates a secure overlay network allowing access to workspaces regardless of network topology, firewalls, or NAT configurations.
-
-### Tailnet and DERP System
-
-The networking system has three key components:
-
-1. **Tailnet**: An overlay network implemented in the `tailnet` package that provides secure, end-to-end encrypted connections between clients, the Coder server, and workspace agents.
-
-2. **DERP Servers**: These relay traffic when direct connections aren't possible. Coder provides several options:
-   - A built-in DERP server that runs on the Coder control plane
-   - Integration with Tailscale's global DERP infrastructure
-   - Support for custom DERP servers for lower latency or offline deployments
-
-3. **Direct Connections**: When possible, the system establishes peer-to-peer connections between clients and workspaces using STUN for NAT traversal. This requires both endpoints to send UDP traffic on ephemeral ports.
-
-### Workspace Proxies
-
-Workspace proxies (in the Enterprise edition) provide regional relay points for browser-based connections, reducing latency for geo-distributed teams. Key characteristics:
-
-- Deployed as independent servers that authenticate with the Coder control plane
-- Relay connections for SSH, workspace apps, port forwarding, and web terminals
-- Do not make direct database connections
-- Managed through the `coder wsproxy` commands
-- Implemented primarily in the `enterprise/wsproxy/` package
-
-## Agent System
-
-The workspace agent runs within each provisioned workspace and provides core functionality including:
-
-- SSH access to workspaces via the `agentssh` package
-- Port forwarding
-- Terminal connectivity via the `pty` package for pseudo-terminal support
-- Application serving
-- Healthcheck monitoring
-- Resource usage reporting
-
-Agents communicate with the control plane using the tailnet system and authenticate using secure tokens.
-
-## Workspace Applications
-
-Workspace applications (or "apps") provide browser-based access to services running within workspaces. The system supports:
-
-- HTTP(S) and WebSocket connections
-- Path-based or subdomain-based access URLs
-- Health checks to monitor application availability
-- Different sharing levels (owner-only, authenticated users, or public)
-- Custom icons and display settings
-
-The implementation is primarily in the `coderd/workspaceapps/` directory with components for URL generation, proxying connections, and managing application state.
-
-## Implementation Details
-
-The project structure separates frontend and backend concerns. React components and pages are organized in the `site/src/` directory, with Jest used for testing. The backend is primarily written in Go, with a strong emphasis on error handling patterns and test coverage.
-
-Database interactions are carefully managed through migrations in `coderd/database/migrations/` and queries in `coderd/database/queries/`. All new queries require proper database authorization (dbauthz) implementation to ensure that only users with appropriate permissions can access specific resources.
-
-## Authorization System
-
-The database authorization (dbauthz) system enforces fine-grained access control across all database operations. It uses role-based access control (RBAC) to validate user permissions before executing database operations. The `dbauthz` package wraps the database store and performs authorization checks before returning data. All database operations must pass through this layer to ensure security.
-
-## Testing Framework
-
-The codebase has a comprehensive testing approach with several key components:
-
-1. **Parallel Testing**: All tests must use `t.Parallel()` to run concurrently, which improves test suite performance and helps identify race conditions.
-
-2. **coderdtest Package**: This package in `coderd/coderdtest/` provides utilities for creating test instances of the Coder server, setting up test users and workspaces, and mocking external components.
-
-3. **Integration Tests**: Tests often span multiple components to verify system behavior, such as template creation, workspace provisioning, and agent connectivity.
-
-4. **Enterprise Testing**: Enterprise features have dedicated test utilities in the `coderdenttest` package.
-
-## Open Source and Enterprise Components
-
-The repository contains both open source and enterprise components:
-
-- Enterprise code lives primarily in the `enterprise/` directory
-- Enterprise features focus on governance, scalability (high availability), and advanced deployment options like workspace proxies
-- The boundary between open source and enterprise is managed through a licensing system
-- The same core codebase supports both editions, with enterprise features conditionally enabled
-
-## Development Philosophy
-
-Coder emphasizes clear error handling, with specific patterns required:
-
-- Concise error messages that avoid phrases like "failed to"
-- Wrapping errors with `%w` to maintain error chains
-- Using sentinel errors with the "err" prefix (e.g., `errNotFound`)
-
-All tests should run in parallel using `t.Parallel()` to ensure efficient testing and expose potential race conditions. The codebase is rigorously linted with golangci-lint to maintain consistent code quality.
-
-Git contributions follow a standard format with commit messages structured as `type: <message>`, where type is one of `feat`, `fix`, or `chore`.
-
-## Development Workflow
-
-Development can be initiated using `scripts/develop.sh` to start the application after making changes. Database schema updates should be performed through the migration system using `create_migration.sh <name>` to generate migration files, with each `.up.sql` migration paired with a corresponding `.down.sql` that properly reverts all changes.
-
-If the development database gets into a bad state, it can be completely reset by removing the PostgreSQL data directory with `rm -rf .coderv2/postgres`. This will destroy all data in the development database, requiring you to recreate any test users, templates, or workspaces after restarting the application.
-
-Code generation for the database layer uses `coderd/database/generate.sh`, and developers should refer to `sqlc.yaml` for the appropriate style and patterns to follow when creating new queries or tables.
-
-The focus should always be on maintaining security through proper database authorization, clean error handling, and comprehensive test coverage to ensure the platform remains robust and reliable.

.cursorrules

@@ -0,0 +1 @@
+AGENTS.md

.devcontainer/devcontainer.json

@@ -1,11 +1,16 @@
 {
 	"name": "Development environments on your infrastructure",
 	"image": "codercom/oss-dogfood:latest",
-
 	"features": {
-		// See all possible options here https://github.com/devcontainers/features/tree/main/src/docker-in-docker
 		"ghcr.io/devcontainers/features/docker-in-docker:2": {
 			"moby": "false"
+		},
+		"ghcr.io/coder/devcontainer-features/code-server:1": {
+			"auth": "none",
+			"port": 13337
+		},
+		"./filebrowser": {
+			"folder": "${containerWorkspaceFolder}"
 		}
 	},
 	// SYS_PTRACE to enable go debugging
@@ -13,6 +18,65 @@
 	"customizations": {
 		"vscode": {
 			"extensions": ["biomejs.biome"]
+		},
+		"coder": {
+			"apps": [
+				{
+					"slug": "cursor",
+					"displayName": "Cursor Desktop",
+					"url": "cursor://coder.coder-remote/openDevContainer?owner=${localEnv:CODER_WORKSPACE_OWNER_NAME}&workspace=${localEnv:CODER_WORKSPACE_NAME}&agent=${localEnv:CODER_WORKSPACE_PARENT_AGENT_NAME}&url=${localEnv:CODER_URL}&token=$SESSION_TOKEN&devContainerName=${localEnv:CONTAINER_ID}&devContainerFolder=${containerWorkspaceFolder}&localWorkspaceFolder=${localWorkspaceFolder}",
+					"external": true,
+					"icon": "/icon/cursor.svg",
+					"order": 1
+				},
+				{
+					"slug": "windsurf",
+					"displayName": "Windsurf Editor",
+					"url": "windsurf://coder.coder-remote/openDevContainer?owner=${localEnv:CODER_WORKSPACE_OWNER_NAME}&workspace=${localEnv:CODER_WORKSPACE_NAME}&agent=${localEnv:CODER_WORKSPACE_PARENT_AGENT_NAME}&url=${localEnv:CODER_URL}&token=$SESSION_TOKEN&devContainerName=${localEnv:CONTAINER_ID}&devContainerFolder=${containerWorkspaceFolder}&localWorkspaceFolder=${localWorkspaceFolder}",
+					"external": true,
+					"icon": "/icon/windsurf.svg",
+					"order": 4
+				},
+				{
+					"slug": "zed",
+					"displayName": "Zed Editor",
+					"url": "zed://ssh/${localEnv:CODER_WORKSPACE_AGENT_NAME}.${localEnv:CODER_WORKSPACE_NAME}.${localEnv:CODER_WORKSPACE_OWNER_NAME}.coder${containerWorkspaceFolder}",
+					"external": true,
+					"icon": "/icon/zed.svg",
+					"order": 5
+				},
+				// Reproduce `code-server` app here from the code-server
+				// feature so that we can set the correct folder and order.
+				// Currently, the order cannot be specified via option because
+				// we parse it as a number whereas variable interpolation
+				// results in a string. Additionally we set health check which
+				// is not yet set in the feature.
+				{
+					"slug": "code-server",
+					"displayName": "code-server",
+					"url": "http://${localEnv:FEATURE_CODE_SERVER_OPTION_HOST:127.0.0.1}:${localEnv:FEATURE_CODE_SERVER_OPTION_PORT:8080}/?folder=${containerWorkspaceFolder}",
+					"openIn": "${localEnv:FEATURE_CODE_SERVER_OPTION_APPOPENIN:slim-window}",
+					"share": "${localEnv:FEATURE_CODE_SERVER_OPTION_APPSHARE:owner}",
+					"icon": "/icon/code.svg",
+					"group": "${localEnv:FEATURE_CODE_SERVER_OPTION_APPGROUP:Web Editors}",
+					"order": 3,
+					"healthCheck": {
+						"url": "http://${localEnv:FEATURE_CODE_SERVER_OPTION_HOST:127.0.0.1}:${localEnv:FEATURE_CODE_SERVER_OPTION_PORT:8080}/healthz",
+						"interval": 5,
+						"threshold": 2
+					}
+				}
+			]
 		}
-	}
+	},
+	"mounts": [
+		// Add a volume for the Coder home directory to persist shell history,
+		// and speed up dotfiles init and/or personalization.
+		"source=coder-coder-devcontainer-home,target=/home/coder,type=volume",
+		// Mount the entire home because conditional mounts are not supported.
+		// See: https://github.com/devcontainers/spec/issues/132
+		"source=${localEnv:HOME},target=/mnt/home/coder,type=bind,readonly"
+	],
+	"postCreateCommand": ["./.devcontainer/scripts/post_create.sh"],
+	"postStartCommand": ["./.devcontainer/scripts/post_start.sh"]
 }

.devcontainer/filebrowser/devcontainer-feature.json

@@ -0,0 +1,46 @@
+{
+	"id": "filebrowser",
+	"version": "0.0.1",
+	"name": "File Browser",
+	"description": "A web-based file browser for your development container",
+	"options": {
+		"port": {
+			"type": "string",
+			"default": "13339",
+			"description": "The port to run filebrowser on"
+		},
+		"folder": {
+			"type": "string",
+			"default": "",
+			"description": "The root directory for filebrowser to serve"
+		},
+		"baseUrl": {
+			"type": "string",
+			"default": "",
+			"description": "The base URL for filebrowser (e.g., /filebrowser)"
+		}
+	},
+	"entrypoint": "/usr/local/bin/filebrowser-entrypoint",
+	"dependsOn": {
+		"ghcr.io/devcontainers/features/common-utils:2": {}
+	},
+	"customizations": {
+		"coder": {
+			"apps": [
+				{
+					"slug": "filebrowser",
+					"displayName": "File Browser",
+					"url": "http://localhost:${localEnv:FEATURE_FILEBROWSER_OPTION_PORT:13339}",
+					"icon": "/icon/filebrowser.svg",
+					"order": 3,
+					"subdomain": true,
+					"healthcheck": {
+						"url": "http://localhost:${localEnv:FEATURE_FILEBROWSER_OPTION_PORT:13339}/health",
+						"interval": 5,
+						"threshold": 2
+					}
+				}
+			]
+		}
+	}
+}

.devcontainer/filebrowser/install.sh

@@ -0,0 +1,54 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+BOLD='\033[0;1m'
+
+printf "%sInstalling filebrowser\n\n" "${BOLD}"
+
+# Check if filebrowser is installed.
+if ! command -v filebrowser &>/dev/null; then
+	VERSION="v2.42.1"
+	EXPECTED_HASH="7d83c0f077df10a8ec9bfd9bf6e745da5d172c3c768a322b0e50583a6bc1d3cc"
+
+	curl -fsSL "https://github.com/filebrowser/filebrowser/releases/download/${VERSION}/linux-amd64-filebrowser.tar.gz" -o /tmp/filebrowser.tar.gz
+	echo "${EXPECTED_HASH} /tmp/filebrowser.tar.gz" | sha256sum -c
+	tar -xzf /tmp/filebrowser.tar.gz -C /tmp
+	sudo mv /tmp/filebrowser /usr/local/bin/
+	sudo chmod +x /usr/local/bin/filebrowser
+	rm /tmp/filebrowser.tar.gz
+fi
+
+# Create entrypoint.
+cat >/usr/local/bin/filebrowser-entrypoint <<EOF
+#!/usr/bin/env bash
+
+PORT="${PORT}"
+FOLDER="${FOLDER:-}"
+FOLDER="\${FOLDER:-\$(pwd)}"
+BASEURL="${BASEURL:-}"
+LOG_PATH=/tmp/filebrowser.log
+export FB_DATABASE="\${HOME}/.filebrowser.db"
+
+printf "🛠️ Configuring filebrowser\n\n"
+
+# Check if filebrowser db exists.
+if [[ ! -f "\${FB_DATABASE}" ]]; then
+	filebrowser config init >>\${LOG_PATH} 2>&1
+	filebrowser users add admin "" --perm.admin=true --viewMode=mosaic >>\${LOG_PATH} 2>&1
+fi
+
+filebrowser config set --baseurl=\${BASEURL} --port=\${PORT} --auth.method=noauth --root=\${FOLDER} >>\${LOG_PATH} 2>&1
+
+printf "👷 Starting filebrowser...\n\n"
+
+printf "📂 Serving \${FOLDER} at http://localhost:\${PORT}\n\n"
+
+filebrowser >>\${LOG_PATH} 2>&1 &
+
+printf "📝 Logs at \${LOG_PATH}\n\n"
+EOF
+
+chmod +x /usr/local/bin/filebrowser-entrypoint
+
+printf "🥳 Installation complete!\n\n"

.devcontainer/scripts/post_create.sh

@@ -0,0 +1,67 @@
+#!/bin/sh
+
+install_devcontainer_cli() {
+	set -e
+	echo "🔧 Installing DevContainer CLI..."
+	cd "$(dirname "$0")/../tools/devcontainer-cli"
+	npm ci --omit=dev
+	ln -sf "$(pwd)/node_modules/.bin/devcontainer" "$(npm config get prefix)/bin/devcontainer"
+}
+
+install_ssh_config() {
+	echo "🔑 Installing SSH configuration..."
+	if [ -d /mnt/home/coder/.ssh ]; then
+		rsync -a /mnt/home/coder/.ssh/ ~/.ssh/
+		chmod 0700 ~/.ssh
+	else
+		echo "⚠️ SSH directory not found."
+	fi
+}
+
+install_git_config() {
+	echo "📂 Installing Git configuration..."
+	if [ -f /mnt/home/coder/git/config ]; then
+		rsync -a /mnt/home/coder/git/ ~/.config/git/
+	elif [ -d /mnt/home/coder/.gitconfig ]; then
+		rsync -a /mnt/home/coder/.gitconfig ~/.gitconfig
+	else
+		echo "⚠️ Git configuration directory not found."
+	fi
+}
+
+install_dotfiles() {
+	if [ ! -d /mnt/home/coder/.config/coderv2/dotfiles ]; then
+		echo "⚠️ Dotfiles directory not found."
+		return
+	fi
+
+	cd /mnt/home/coder/.config/coderv2/dotfiles || return
+	for script in install.sh install bootstrap.sh bootstrap script/bootstrap setup.sh setup script/setup; do
+		if [ -x $script ]; then
+			echo "📦 Installing dotfiles..."
+			./$script || {
+				echo "❌ Error running $script. Please check the script for issues."
+				return
+			}
+			echo "✅ Dotfiles installed successfully."
+			return
+		fi
+	done
+	echo "⚠️ No install script found in dotfiles directory."
+}
+
+personalize() {
+	# Allow script to continue as Coder dogfood utilizes a hack to
+	# synchronize startup script execution.
+	touch /tmp/.coder-startup-script.done
+
+	if [ -x /mnt/home/coder/personalize ]; then
+		echo "🎨 Personalizing environment..."
+		/mnt/home/coder/personalize
+	fi
+}
+
+install_devcontainer_cli
+install_ssh_config
+install_dotfiles
+personalize

.devcontainer/scripts/post_start.sh

@@ -0,0 +1,4 @@
+#!/bin/sh
+
+# Start Docker service if not already running.
+sudo service docker start

.devcontainer/tools/devcontainer-cli/package-lock.json

@@ -0,0 +1,26 @@
+{
+  "name": "devcontainer-cli",
+  "version": "1.0.0",
+  "lockfileVersion": 3,
+  "requires": true,
+  "packages": {
+    "": {
+      "name": "devcontainer-cli",
+      "version": "1.0.0",
+      "dependencies": {
+        "@devcontainers/cli": "^0.80.0"
+      }
+    },
+    "node_modules/@devcontainers/cli": {
+      "version": "0.80.0",
+      "resolved": "https://registry.npmjs.org/@devcontainers/cli/-/cli-0.80.0.tgz",
+      "integrity": "sha512-w2EaxgjyeVGyzfA/KUEZBhyXqu/5PyWNXcnrXsZOBrt3aN2zyGiHrXoG54TF6K0b5DSCF01Rt5fnIyrCeFzFKw==",
+      "bin": {
+        "devcontainer": "devcontainer.js"
+      },
+      "engines": {
+        "node": "^16.13.0 || >=18.0.0"
+      }
+    }
+  }
+}

.devcontainer/tools/devcontainer-cli/package.json

@@ -0,0 +1,8 @@
+{
+  "name": "devcontainer-cli",
+  "private": true,
+  "version": "1.0.0",
+  "dependencies": {
+    "@devcontainers/cli": "^0.80.0"
+  }
+}

.editorconfig

@@ -7,7 +7,7 @@ trim_trailing_whitespace = true
 insert_final_newline = true
 indent_style = tab
 
-[*.{yaml,yml,tf,tfvars,nix}]
+[*.{yaml,yml,tf,tftpl,tfvars,nix}]
 indent_style = space
 indent_size = 2
 
@@ -18,3 +18,11 @@ indent_size = 2
 [coderd/database/dump.sql]
 indent_style = space
 indent_size = 4
+
+[coderd/database/queries/*.sql]
+indent_style = tab
+indent_size = 4
+
+[coderd/database/migrations/*.sql]
+indent_style = tab
+indent_size = 4

.gitattributes

@@ -15,6 +15,8 @@ provisionersdk/proto/*.go linguist-generated=true
 *.tfstate.json linguist-generated=true
 *.tfstate.dot linguist-generated=true
 *.tfplan.dot linguist-generated=true
+site/e2e/google/protobuf/timestampGenerated.ts
 site/e2e/provisionerGenerated.ts linguist-generated=true
+site/src/api/countriesGenerated.tsx linguist-generated=true
+site/src/api/rbacresourcesGenerated.tsx linguist-generated=true
 site/src/api/typesGenerated.ts linguist-generated=true
-site/src/pages/SetupPage/countries.tsx linguist-generated=true

.github/.linkspector.yml

@@ -25,5 +25,9 @@ ignorePatterns:
   - pattern: "docs.github.com"
   - pattern: "claude.ai"
   - pattern: "splunk.com"
+  - pattern: "stackoverflow.com/questions"
+  - pattern: "developer.hashicorp.com/terraform/language"
+  - pattern: "platform.openai.com"
+  - pattern: "api.openai.com"
 aliveStatusCodes:
   - 200

.github/actions/embedded-pg-cache/download/action.yml

@@ -0,0 +1,49 @@
+name: "Download Embedded Postgres Cache"
+description: |
+  Downloads the embedded postgres cache and outputs today's cache key.
+  A PR job can use a cache if it was created by its base branch, its current
+  branch, or the default branch.
+  https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/caching-dependencies-to-speed-up-workflows#restrictions-for-accessing-a-cache
+outputs:
+  cache-key:
+    description: "Today's cache key"
+    value: ${{ steps.vars.outputs.cache-key }}
+inputs:
+  key-prefix:
+    description: "Prefix for the cache key"
+    required: true
+  cache-path:
+    description: "Path to the cache directory"
+    required: true
+runs:
+  using: "composite"
+  steps:
+    - name: Get date values and cache key
+      id: vars
+      shell: bash
+      run: |
+        export YEAR_MONTH=$(date +'%Y-%m')
+        export PREV_YEAR_MONTH=$(date -d 'last month' +'%Y-%m')
+        export DAY=$(date +'%d')
+        echo "year-month=$YEAR_MONTH" >> "$GITHUB_OUTPUT"
+        echo "prev-year-month=$PREV_YEAR_MONTH" >> "$GITHUB_OUTPUT"
+        echo "cache-key=${INPUTS_KEY_PREFIX}-${YEAR_MONTH}-${DAY}" >> "$GITHUB_OUTPUT"
+      env:
+        INPUTS_KEY_PREFIX: ${{ inputs.key-prefix }}
+
+    # By default, depot keeps caches for 14 days. This is plenty for embedded
+    # postgres, which changes infrequently.
+    # https://depot.dev/docs/github-actions/overview#cache-retention-policy
+    - name: Download embedded Postgres cache
+      uses: actions/cache/restore@5a3ec84eff668545956fd18022155c47e93e2684 # v4.2.3
+      with:
+        path: ${{ inputs.cache-path }}
+        key: ${{ steps.vars.outputs.cache-key }}
+        # > If there are multiple partial matches for a restore key, the action returns the most recently created cache.
+        # https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/caching-dependencies-to-speed-up-workflows#matching-a-cache-key
+        # The second restore key allows non-main branches to use the cache from the previous month.
+        # This prevents PRs from rebuilding the cache on the first day of the month.
+        # It also makes sure that once a month, the cache is fully reset.
+        restore-keys: |
+          ${{ inputs.key-prefix }}-${{ steps.vars.outputs.year-month }}-
+          ${{ github.ref != 'refs/heads/main' && format('{0}-{1}-', inputs.key-prefix, steps.vars.outputs.prev-year-month) || '' }}

.github/actions/embedded-pg-cache/upload/action.yml

@@ -0,0 +1,18 @@
+name: "Upload Embedded Postgres Cache"
+description: Uploads the embedded Postgres cache. This only runs on the main branch.
+inputs:
+  cache-key:
+    description: "Cache key"
+    required: true
+  cache-path:
+    description: "Path to the cache directory"
+    required: true
+runs:
+  using: "composite"
+  steps:
+    - name: Upload Embedded Postgres cache
+      if: ${{ github.ref == 'refs/heads/main' }}
+      uses: actions/cache/save@5a3ec84eff668545956fd18022155c47e93e2684 # v4.2.3
+      with:
+        path: ${{ inputs.cache-path }}
+        key: ${{ inputs.cache-key }}

.github/actions/setup-embedded-pg-cache-paths/action.yml

@@ -0,0 +1,33 @@
+name: "Setup Embedded Postgres Cache Paths"
+description: Sets up a path for cached embedded postgres binaries.
+outputs:
+  embedded-pg-cache:
+    description: "Value of EMBEDDED_PG_CACHE_DIR"
+    value: ${{ steps.paths.outputs.embedded-pg-cache }}
+  cached-dirs:
+    description: "directories that should be cached between CI runs"
+    value: ${{ steps.paths.outputs.cached-dirs }}
+runs:
+  using: "composite"
+  steps:
+    - name: Override Go paths
+      id: paths
+      uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea # v7
+      with:
+        script: |
+          const path = require('path');
+
+          // RUNNER_TEMP should be backed by a RAM disk on Windows if
+          // coder/setup-ramdisk-action was used
+          const runnerTemp = process.env.RUNNER_TEMP;
+          const embeddedPgCacheDir = path.join(runnerTemp, 'embedded-pg-cache');
+          core.exportVariable('EMBEDDED_PG_CACHE_DIR', embeddedPgCacheDir);
+          core.setOutput('embedded-pg-cache', embeddedPgCacheDir);
+          const cachedDirs = `${embeddedPgCacheDir}`;
+          core.setOutput('cached-dirs', cachedDirs);
+
+    - name: Create directories
+      shell: bash
+      run: |
+        set -e
+        mkdir -p "$EMBEDDED_PG_CACHE_DIR"

.github/actions/setup-go/action.yaml

@@ -4,7 +4,7 @@ description: |
 inputs:
   version:
     description: "The Go version to use."
-    default: "1.24.6"
+    default: "1.24.10"
   use-preinstalled-go:
     description: "Whether to use preinstalled Go."
     default: "false"

.github/actions/setup-node/action.yaml

@@ -16,7 +16,7 @@ runs:
     - name: Setup Node
       uses: actions/setup-node@0a44ba7841725637a19e28fa30b79a866c81b0a6 # v4.0.4
       with:
-        node-version: 20.16.0
+        node-version: 22.19.0
         # See https://github.com/actions/setup-node#caching-global-packages-data
         cache: "pnpm"
         cache-dependency-path: ${{ inputs.directory }}/pnpm-lock.yaml

.github/actions/setup-sqlc/action.yaml

@@ -5,6 +5,13 @@ runs:
   using: "composite"
   steps:
     - name: Setup sqlc
-      uses: sqlc-dev/setup-sqlc@c0209b9199cd1cce6a14fc27cabcec491b651761 # v4.0.0
-      with:
-        sqlc-version: "1.27.0"
+      # uses: sqlc-dev/setup-sqlc@c0209b9199cd1cce6a14fc27cabcec491b651761 # v4.0.0
+      # with:
+      #   sqlc-version: "1.30.0"
+
+      # Switched to coder/sqlc fork to fix ambiguous column bug, see:
+      # - https://github.com/coder/sqlc/pull/1
+      # - https://github.com/sqlc-dev/sqlc/pull/4159
+      shell: bash
+      run: |
+        CGO_ENABLED=1 go install github.com/coder/sqlc/cmd/sqlc@aab4e865a51df0c43e1839f81a9d349b41d14f05

.github/actions/setup-tf/action.yaml

@@ -7,5 +7,5 @@ runs:
     - name: Install Terraform
       uses: hashicorp/setup-terraform@b9cd54a3c349d3f38e8881555d616ced269862dd # v3.1.2
       with:
-        terraform_version: 1.12.2
+        terraform_version: 1.13.4
         terraform_wrapper: false

.github/actions/test-cache/download/action.yml

@@ -27,9 +27,11 @@ runs:
         export YEAR_MONTH=$(date +'%Y-%m')
         export PREV_YEAR_MONTH=$(date -d 'last month' +'%Y-%m')
         export DAY=$(date +'%d')
-        echo "year-month=$YEAR_MONTH" >> $GITHUB_OUTPUT
-        echo "prev-year-month=$PREV_YEAR_MONTH" >> $GITHUB_OUTPUT
-        echo "cache-key=${{ inputs.key-prefix }}-${YEAR_MONTH}-${DAY}" >> $GITHUB_OUTPUT
+        echo "year-month=$YEAR_MONTH" >> "$GITHUB_OUTPUT"
+        echo "prev-year-month=$PREV_YEAR_MONTH" >> "$GITHUB_OUTPUT"
+        echo "cache-key=${INPUTS_KEY_PREFIX}-${YEAR_MONTH}-${DAY}" >> "$GITHUB_OUTPUT"
+      env:
+        INPUTS_KEY_PREFIX: ${{ inputs.key-prefix }}
 
     # TODO: As a cost optimization, we could remove caches that are older than
     # a day or two. By default, depot keeps caches for 14 days, which isn't

.github/actions/upload-datadog/action.yaml

@@ -12,13 +12,12 @@ runs:
       run: |
         set -e
 
-        owner=${{ github.repository_owner	 }}
-        echo "owner: $owner"
-        if [[  $owner != "coder" ]]; then
+        echo "owner: $REPO_OWNER"
+        if [[ "$REPO_OWNER" != "coder" ]]; then
           echo "Not a pull request from the main repo, skipping..."
           exit 0
         fi
-        if [[ -z "${{ inputs.api-key }}" ]]; then
+        if [[ -z "${DATADOG_API_KEY}" ]]; then
           # This can happen for dependabot.
           echo "No API key provided, skipping..."
           exit 0
@@ -31,37 +30,38 @@ runs:
 
         TMP_DIR=$(mktemp -d)
 
-        if [[ "${{ runner.os }}" == "Windows" ]]; then
+        if [[ "${RUNNER_OS}" == "Windows" ]]; then
           BINARY_PATH="${TMP_DIR}/datadog-ci.exe"
           BINARY_URL="https://github.com/DataDog/datadog-ci/releases/download/${BINARY_VERSION}/datadog-ci_win-x64"
-        elif [[ "${{ runner.os }}" == "macOS" ]]; then
+        elif [[ "${RUNNER_OS}" == "macOS" ]]; then
           BINARY_PATH="${TMP_DIR}/datadog-ci"
           BINARY_URL="https://github.com/DataDog/datadog-ci/releases/download/${BINARY_VERSION}/datadog-ci_darwin-arm64"
-        elif [[ "${{ runner.os }}" == "Linux" ]]; then
+        elif [[ "${RUNNER_OS}" == "Linux" ]]; then
           BINARY_PATH="${TMP_DIR}/datadog-ci"
           BINARY_URL="https://github.com/DataDog/datadog-ci/releases/download/${BINARY_VERSION}/datadog-ci_linux-x64"
         else
-          echo "Unsupported OS: ${{ runner.os }}"
+          echo "Unsupported OS: $RUNNER_OS"
           exit 1
         fi
 
-        echo "Downloading DataDog CI binary version ${BINARY_VERSION} for ${{ runner.os }}..."
+        echo "Downloading DataDog CI binary version ${BINARY_VERSION} for $RUNNER_OS..."
         curl -sSL "$BINARY_URL" -o "$BINARY_PATH"
 
-        if [[ "${{ runner.os }}" == "Windows" ]]; then
+        if [[ "${RUNNER_OS}" == "Windows" ]]; then
           echo "$BINARY_HASH_WINDOWS  $BINARY_PATH" | sha256sum --check
-        elif [[ "${{ runner.os }}" == "macOS" ]]; then
+        elif [[ "${RUNNER_OS}" == "macOS" ]]; then
           echo "$BINARY_HASH_MACOS  $BINARY_PATH" | shasum -a 256 --check
-        elif [[ "${{ runner.os }}" == "Linux" ]]; then
+        elif [[ "${RUNNER_OS}" == "Linux" ]]; then
           echo "$BINARY_HASH_LINUX  $BINARY_PATH" | sha256sum --check
         fi
 
         # Make binary executable (not needed for Windows)
-        if [[ "${{ runner.os }}" != "Windows" ]]; then
+        if [[ "${RUNNER_OS}" != "Windows" ]]; then
           chmod +x "$BINARY_PATH"
         fi
 
         "$BINARY_PATH" junit upload --service coder ./gotests.xml \
-          --tags os:${{runner.os}} --tags runner_name:${{runner.name}}
+          --tags "os:${RUNNER_OS}" --tags "runner_name:${RUNNER_NAME}"
       env:
+        REPO_OWNER: ${{ github.repository_owner }}
         DATADOG_API_KEY: ${{ inputs.api-key }}

.github/dependabot.yaml

@@ -33,6 +33,7 @@ updates:
       - dependency-name: "*"
         update-types:
           - version-update:semver-patch
+      - dependency-name: "github.com/mark3labs/mcp-go"
 
   # Update our Dockerfile.
   - package-ecosystem: "docker"
@@ -79,6 +80,9 @@ updates:
       mui:
         patterns:
           - "@mui*"
+      radix:
+        patterns:
+          - "@radix-ui/*"
       react:
         patterns:
           - "react"
@@ -103,6 +107,7 @@ updates:
       - dependency-name: "*"
         update-types:
           - version-update:semver-major
+      - dependency-name: "@playwright/test"
     open-pull-requests-limit: 15
 
   - package-ecosystem: "terraform"
@@ -114,9 +119,9 @@ updates:
     commit-message:
       prefix: "chore"
     groups:
-      coder:
+      coder-modules:
         patterns:
-          - "registry.coder.com/coder/*/coder"
+          - "coder/*/coder"
     labels: []
     ignore:
       - dependency-name: "*"

.github/fly-wsproxies/sao-paulo-coder.toml

@@ -1,34 +0,0 @@
-app = "sao-paulo-coder"
-primary_region = "gru"
-
-[experimental]
-  entrypoint = ["/bin/sh", "-c", "CODER_DERP_SERVER_RELAY_URL=\"http://[${FLY_PRIVATE_IP}]:3000\" /opt/coder wsproxy server"]
-  auto_rollback = true
-
-[build]
-  image = "ghcr.io/coder/coder-preview:main"
-
-[env]
-  CODER_ACCESS_URL = "https://sao-paulo.fly.dev.coder.com"
-  CODER_HTTP_ADDRESS = "0.0.0.0:3000"
-  CODER_PRIMARY_ACCESS_URL = "https://dev.coder.com"
-  CODER_WILDCARD_ACCESS_URL = "*--apps.sao-paulo.fly.dev.coder.com"
-  CODER_VERBOSE = "true"
-
-[http_service]
-  internal_port = 3000
-  force_https = true
-  auto_stop_machines = true
-  auto_start_machines = true
-  min_machines_running = 0
-
-# Ref: https://fly.io/docs/reference/configuration/#http_service-concurrency
-[http_service.concurrency]
-  type = "requests"
-  soft_limit = 50
-  hard_limit = 100
-
-[[vm]]
-  cpu_kind = "shared"
-  cpus = 2
-  memory_mb = 512

.github/pull_request_template.md

@@ -0,0 +1,5 @@
+<!--
+
+If you have used AI to produce some or all of this PR, please ensure you have read our [AI Contribution guidelines](https://coder.com/docs/about/contributing/AI_CONTRIBUTING) before submitting.
+
+-->

.github/workflows/contrib.yaml

@@ -3,6 +3,7 @@ name: contrib
 on:
   issue_comment:
     types: [created, edited]
+  # zizmor: ignore[dangerous-triggers] We explicitly want to run on pull_request_target.
   pull_request_target:
     types:
       - opened
@@ -52,7 +53,7 @@ jobs:
     if: ${{ github.event_name == 'pull_request_target' && !github.event.pull_request.draft }}
     steps:
       - name: release-labels
-        uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea # v7.0.1
+        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
         with:
           # This script ensures PR title and labels are in sync:
           #

.github/workflows/dependabot.yaml

@@ -15,7 +15,7 @@ jobs:
       github.event_name == 'pull_request' &&
       github.event.action == 'opened' &&
       github.event.pull_request.user.login == 'dependabot[bot]' &&
-      github.actor_id == 49699333 &&
+      github.event.pull_request.user.id == 49699333 &&
       github.repository == 'coder/coder'
     permissions:
       pull-requests: write
@@ -28,6 +28,7 @@ jobs:
           github-token: "${{ secrets.GITHUB_TOKEN }}"
 
       - name: Approve the PR
+        if: steps.metadata.outputs.package-ecosystem != 'github-actions'
         run: |
           echo "Approving $PR_URL"
           gh pr review --approve "$PR_URL"
@@ -36,6 +37,7 @@ jobs:
           GH_TOKEN: ${{secrets.GITHUB_TOKEN}}
 
       - name: Enable auto-merge
+        if: steps.metadata.outputs.package-ecosystem != 'github-actions'
         run: |
           echo "Enabling auto-merge for $PR_URL"
           gh pr merge --auto --squash "$PR_URL"
@@ -44,11 +46,12 @@ jobs:
           GH_TOKEN: ${{secrets.GITHUB_TOKEN}}
 
       - name: Send Slack notification
-        env:
-          PR_URL: ${{github.event.pull_request.html_url}}
-          PR_TITLE: ${{github.event.pull_request.title}}
-          PR_NUMBER: ${{github.event.pull_request.number}}
         run: |
+          if [ "$PACKAGE_ECOSYSTEM" = "github-actions" ]; then
+            STATUS_TEXT=":pr-opened: Dependabot opened PR #${PR_NUMBER} (GitHub Actions changes are not auto-merged)"
+          else
+            STATUS_TEXT=":pr-merged: Auto merge enabled for Dependabot PR #${PR_NUMBER}"
+          fi
           curl -X POST -H 'Content-type: application/json' \
           --data '{
             "username": "dependabot",
@@ -58,7 +61,7 @@ jobs:
                 "type": "header",
                 "text": {
                   "type": "plain_text",
-                  "text": ":pr-merged: Auto merge enabled for Dependabot PR #${{ env.PR_NUMBER }}",
+                  "text": "'"${STATUS_TEXT}"'",
                   "emoji": true
                 }
               },
@@ -67,7 +70,7 @@ jobs:
                 "fields": [
                   {
                     "type": "mrkdwn",
-                    "text": "${{ env.PR_TITLE }}"
+                    "text": "'"${PR_TITLE}"'"
                   }
                 ]
               },
@@ -80,9 +83,15 @@ jobs:
                       "type": "plain_text",
                       "text": "View PR"
                     },
-                    "url": "${{ env.PR_URL }}"
+                    "url": "'"${PR_URL}"'"
                   }
                 ]
               }
             ]
-          }' ${{ secrets.DEPENDABOT_PRS_SLACK_WEBHOOK }}
+          }' "${{ secrets.DEPENDABOT_PRS_SLACK_WEBHOOK }}"
+        env:
+          SLACK_WEBHOOK: ${{ secrets.DEPENDABOT_PRS_SLACK_WEBHOOK }}
+          PACKAGE_ECOSYSTEM: ${{ steps.metadata.outputs.package-ecosystem }}
+          PR_NUMBER: ${{ github.event.pull_request.number }}
+          PR_TITLE: ${{ github.event.pull_request.title }}
+          PR_URL: ${{ github.event.pull_request.html_url }}

.github/workflows/deploy.yaml

@@ -0,0 +1,172 @@
+name: deploy
+
+on:
+  # Via workflow_call, called from ci.yaml
+  workflow_call:
+    inputs:
+      image:
+        description: "Image and tag to potentially deploy. Current branch will be validated against should-deploy check."
+        required: true
+        type: string
+    secrets:
+      FLY_API_TOKEN:
+        required: true
+      FLY_PARIS_CODER_PROXY_SESSION_TOKEN:
+        required: true
+      FLY_SYDNEY_CODER_PROXY_SESSION_TOKEN:
+        required: true
+      FLY_SAO_PAULO_CODER_PROXY_SESSION_TOKEN:
+        required: true
+      FLY_JNB_CODER_PROXY_SESSION_TOKEN:
+        required: true
+
+permissions:
+  contents: read
+
+concurrency:
+  group: ${{ github.workflow }} # no per-branch concurrency
+  cancel-in-progress: false
+
+jobs:
+  # Determines if the given branch should be deployed to dogfood.
+  should-deploy:
+    name: should-deploy
+    runs-on: ubuntu-latest
+    outputs:
+      verdict: ${{ steps.check.outputs.verdict }} # DEPLOY or NOOP
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
+        with:
+          egress-policy: audit
+
+      - name: Checkout
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          fetch-depth: 0
+          persist-credentials: false
+
+      - name: Check if deploy is enabled
+        id: check
+        run: |
+          set -euo pipefail
+          verdict="$(./scripts/should_deploy.sh)"
+          echo "verdict=$verdict" >> "$GITHUB_OUTPUT"
+
+  deploy:
+    name: "deploy"
+    runs-on: ubuntu-latest
+    timeout-minutes: 30
+    needs: should-deploy
+    if: needs.should-deploy.outputs.verdict == 'DEPLOY'
+    permissions:
+      contents: read
+      id-token: write
+      packages: write # to retag image as dogfood
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
+        with:
+          egress-policy: audit
+
+      - name: Checkout
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          fetch-depth: 0
+          persist-credentials: false
+
+      - name: GHCR Login
+        uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # v3.6.0
+        with:
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Authenticate to Google Cloud
+        uses: google-github-actions/auth@7c6bc770dae815cd3e89ee6cdf493a5fab2cc093 # v3.0.0
+        with:
+          workload_identity_provider: ${{ vars.GCP_WORKLOAD_ID_PROVIDER }}
+          service_account: ${{ vars.GCP_SERVICE_ACCOUNT }}
+
+      - name: Set up Google Cloud SDK
+        uses: google-github-actions/setup-gcloud@aa5489c8933f4cc7a4f7d45035b3b1440c9c10db # v3.0.1
+
+      - name: Set up Flux CLI
+        uses: fluxcd/flux2/action@8454b02a32e48d775b9f563cb51fdcb1787b5b93 # v2.7.5
+        with:
+          # Keep this and the github action up to date with the version of flux installed in dogfood cluster
+          version: "2.7.0"
+
+      - name: Get Cluster Credentials
+        uses: google-github-actions/get-gke-credentials@3da1e46a907576cefaa90c484278bb5b259dd395 # v3.0.0
+        with:
+          cluster_name: dogfood-v2
+          location: us-central1-a
+          project_id: coder-dogfood-v2
+
+      # Retag image as dogfood while maintaining the multi-arch manifest
+      - name: Tag image as dogfood
+        run: docker buildx imagetools create --tag "ghcr.io/coder/coder-preview:dogfood" "$IMAGE"
+        env:
+          IMAGE: ${{ inputs.image }}
+
+      - name: Reconcile Flux
+        run: |
+          set -euxo pipefail
+          flux --namespace flux-system reconcile source git flux-system
+          flux --namespace flux-system reconcile source git coder-main
+          flux --namespace flux-system reconcile kustomization flux-system
+          flux --namespace flux-system reconcile kustomization coder
+          flux --namespace flux-system reconcile source chart coder-coder
+          flux --namespace flux-system reconcile source chart coder-coder-provisioner
+          flux --namespace coder reconcile helmrelease coder
+          flux --namespace coder reconcile helmrelease coder-provisioner
+          flux --namespace coder reconcile helmrelease coder-provisioner-tagged
+          flux --namespace coder reconcile helmrelease coder-provisioner-tagged-prebuilds
+
+      # Just updating Flux is usually not enough. The Helm release may get
+      # redeployed, but unless something causes the Deployment to update the
+      # pods won't be recreated. It's important that the pods get recreated,
+      # since we use `imagePullPolicy: Always` to ensure we're running the
+      # latest image.
+      - name: Rollout Deployment
+        run: |
+          set -euxo pipefail
+          kubectl --namespace coder rollout restart deployment/coder
+          kubectl --namespace coder rollout status deployment/coder
+          kubectl --namespace coder rollout restart deployment/coder-provisioner
+          kubectl --namespace coder rollout status deployment/coder-provisioner
+          kubectl --namespace coder rollout restart deployment/coder-provisioner-tagged
+          kubectl --namespace coder rollout status deployment/coder-provisioner-tagged
+          kubectl --namespace coder rollout restart deployment/coder-provisioner-tagged-prebuilds
+          kubectl --namespace coder rollout status deployment/coder-provisioner-tagged-prebuilds
+
+  deploy-wsproxies:
+    runs-on: ubuntu-latest
+    needs: deploy
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
+        with:
+          egress-policy: audit
+
+      - name: Checkout
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          fetch-depth: 0
+          persist-credentials: false
+
+      - name: Setup flyctl
+        uses: superfly/flyctl-actions/setup-flyctl@fc53c09e1bc3be6f54706524e3b82c4f462f77be # v1.5
+
+      - name: Deploy workspace proxies
+        run: |
+          flyctl deploy --image "$IMAGE" --app paris-coder --config ./.github/fly-wsproxies/paris-coder.toml --env "CODER_PROXY_SESSION_TOKEN=$TOKEN_PARIS" --yes
+          flyctl deploy --image "$IMAGE" --app sydney-coder --config ./.github/fly-wsproxies/sydney-coder.toml --env "CODER_PROXY_SESSION_TOKEN=$TOKEN_SYDNEY" --yes
+          flyctl deploy --image "$IMAGE" --app jnb-coder --config ./.github/fly-wsproxies/jnb-coder.toml --env "CODER_PROXY_SESSION_TOKEN=$TOKEN_JNB" --yes
+        env:
+          FLY_API_TOKEN: ${{ secrets.FLY_API_TOKEN }}
+          IMAGE: ${{ inputs.image }}
+          TOKEN_PARIS: ${{ secrets.FLY_PARIS_CODER_PROXY_SESSION_TOKEN }}
+          TOKEN_SYDNEY: ${{ secrets.FLY_SYDNEY_CODER_PROXY_SESSION_TOKEN }}
+          TOKEN_JNB: ${{ secrets.FLY_JNB_CODER_PROXY_SESSION_TOKEN }}

.github/workflows/doc-check.yaml

@@ -0,0 +1,205 @@
+# This workflow checks if a PR requires documentation updates.
+# It creates a Coder Task that uses AI to analyze the PR changes,
+# search existing docs, and comment with recommendations.
+#
+# Triggered by: Adding the "doc-check" label to a PR, or manual dispatch.
+
+name: AI Documentation Check
+
+on:
+  pull_request:
+    types:
+      - labeled
+  workflow_dispatch:
+    inputs:
+      pr_url:
+        description: "Pull Request URL to check"
+        required: true
+        type: string
+      template_preset:
+        description: "Template preset to use"
+        required: false
+        default: ""
+        type: string
+
+jobs:
+  doc-check:
+    name: Analyze PR for Documentation Updates Needed
+    runs-on: ubuntu-latest
+    if: |
+      (github.event.label.name == 'doc-check' || github.event_name == 'workflow_dispatch') &&
+      (github.event.pull_request.draft == false || github.event_name == 'workflow_dispatch')
+    timeout-minutes: 30
+    env:
+      CODER_URL: ${{ secrets.DOC_CHECK_CODER_URL }}
+      CODER_SESSION_TOKEN: ${{ secrets.DOC_CHECK_CODER_SESSION_TOKEN }}
+    permissions:
+      contents: read
+      pull-requests: write
+      actions: write
+
+    steps:
+      - name: Determine PR Context
+        id: determine-context
+        env:
+          GITHUB_ACTOR: ${{ github.actor }}
+          GITHUB_EVENT_NAME: ${{ github.event_name }}
+          GITHUB_EVENT_PR_HTML_URL: ${{ github.event.pull_request.html_url }}
+          GITHUB_EVENT_PR_NUMBER: ${{ github.event.pull_request.number }}
+          GITHUB_EVENT_SENDER_ID: ${{ github.event.sender.id }}
+          GITHUB_EVENT_SENDER_LOGIN: ${{ github.event.sender.login }}
+          INPUTS_PR_URL: ${{ inputs.pr_url }}
+          INPUTS_TEMPLATE_PRESET: ${{ inputs.template_preset || '' }}
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          echo "Using template preset: ${INPUTS_TEMPLATE_PRESET}"
+          echo "template_preset=${INPUTS_TEMPLATE_PRESET}" >> "${GITHUB_OUTPUT}"
+
+          # For workflow_dispatch, use the provided PR URL
+          if [[ "${GITHUB_EVENT_NAME}" == "workflow_dispatch" ]]; then
+            if ! GITHUB_USER_ID=$(gh api "users/${GITHUB_ACTOR}" --jq '.id'); then
+              echo "::error::Failed to get GitHub user ID for actor ${GITHUB_ACTOR}"
+              exit 1
+            fi
+            echo "Using workflow_dispatch actor: ${GITHUB_ACTOR} (ID: ${GITHUB_USER_ID})"
+            echo "github_user_id=${GITHUB_USER_ID}" >> "${GITHUB_OUTPUT}"
+            echo "github_username=${GITHUB_ACTOR}" >> "${GITHUB_OUTPUT}"
+
+            echo "Using PR URL: ${INPUTS_PR_URL}"
+            # Convert /pull/ to /issues/ for create-task-action compatibility
+            ISSUE_URL="${INPUTS_PR_URL/\/pull\//\/issues\/}"
+            echo "pr_url=${ISSUE_URL}" >> "${GITHUB_OUTPUT}"
+
+            # Extract PR number from URL for later use
+            PR_NUMBER=$(echo "${INPUTS_PR_URL}" | grep -oP '(?<=pull/)\d+')
+            echo "pr_number=${PR_NUMBER}" >> "${GITHUB_OUTPUT}"
+
+          elif [[ "${GITHUB_EVENT_NAME}" == "pull_request" ]]; then
+            GITHUB_USER_ID=${GITHUB_EVENT_SENDER_ID}
+            echo "Using label adder: ${GITHUB_EVENT_SENDER_LOGIN} (ID: ${GITHUB_USER_ID})"
+            echo "github_user_id=${GITHUB_USER_ID}" >> "${GITHUB_OUTPUT}"
+            echo "github_username=${GITHUB_EVENT_SENDER_LOGIN}" >> "${GITHUB_OUTPUT}"
+
+            echo "Using PR URL: ${GITHUB_EVENT_PR_HTML_URL}"
+            # Convert /pull/ to /issues/ for create-task-action compatibility
+            ISSUE_URL="${GITHUB_EVENT_PR_HTML_URL/\/pull\//\/issues\/}"
+            echo "pr_url=${ISSUE_URL}" >> "${GITHUB_OUTPUT}"
+            echo "pr_number=${GITHUB_EVENT_PR_NUMBER}" >> "${GITHUB_OUTPUT}"
+
+          else
+            echo "::error::Unsupported event type: ${GITHUB_EVENT_NAME}"
+            exit 1
+          fi
+
+      - name: Extract changed files and build prompt
+        id: extract-context
+        env:
+          PR_URL: ${{ steps.determine-context.outputs.pr_url }}
+          PR_NUMBER: ${{ steps.determine-context.outputs.pr_number }}
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          echo "Analyzing PR #${PR_NUMBER}"
+
+          # Build task prompt - using unquoted heredoc so variables expand
+          TASK_PROMPT=$(cat <<EOF
+          Review PR #${PR_NUMBER} and determine if documentation needs updating or creating.
+
+          PR URL: ${PR_URL}
+
+          WORKFLOW:
+            1. Setup (repo is pre-cloned at ~/coder)
+               cd ~/coder
+               git fetch origin pull/${PR_NUMBER}/head:pr-${PR_NUMBER}
+               git checkout pr-${PR_NUMBER}
+
+            2. Get PR info
+               Use GitHub MCP tools to get PR title, body, and diff
+               Or use: git diff main...pr-${PR_NUMBER}
+
+            3. Understand Changes
+               Read the diff and identify what changed
+               Ask: Is this user-facing? Does it change behavior? Is it a new feature?
+
+            4. Search for Related Docs
+               cat ~/coder/docs/manifest.json | jq '.routes[] | {title, path}' | head -50
+               grep -ri "relevant_term" ~/coder/docs/ --include="*.md"
+
+            5. Decide
+               NEEDS DOCS if: New feature, API change, CLI change, behavior change, user-visible
+               NO DOCS if: Internal refactor, test-only, already documented, non-user-facing, dependency updates
+               FIRST check: Did this PR already update docs? If yes and complete, say "No Changes Needed"
+
+            6. Comment on the PR using this format
+
+          COMMENT FORMAT:
+          ## 📚 Documentation Check
+
+          ### ✅ Updates Needed
+          - **[docs/path/file.md](github_link)** - Brief what needs changing
+
+          ### 📝 New Docs Needed
+          - **docs/suggested/location.md** - What should be documented
+
+          ### ✨ No Changes Needed
+          [Reason: Documents already updated in PR | Internal changes only | Test-only | No user-facing impact]
+
+          ---
+          *This comment was generated by an AI Agent through [Coder Tasks](https://coder.com/docs/ai-coder/tasks)*
+
+          DOCS STRUCTURE:
+            Read ~/coder/docs/manifest.json for the complete documentation structure.
+            Common areas include: reference/, admin/, user-guides/, ai-coder/, install/, tutorials/
+            But check manifest.json - it has everything.
+
+          EOF
+          )
+
+          # Output the prompt
+          {
+            echo "task_prompt<<EOFOUTPUT"
+            echo "${TASK_PROMPT}"
+            echo "EOFOUTPUT"
+          } >> "${GITHUB_OUTPUT}"
+
+      - name: Checkout create-task-action
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          fetch-depth: 1
+          path: ./.github/actions/create-task-action
+          persist-credentials: false
+          ref: main
+          repository: coder/create-task-action
+
+      - name: Create Coder Task for Documentation Check
+        id: create_task
+        uses: ./.github/actions/create-task-action
+        with:
+          coder-url: ${{ secrets.DOC_CHECK_CODER_URL }}
+          coder-token: ${{ secrets.DOC_CHECK_CODER_SESSION_TOKEN }}
+          coder-organization: "default"
+          coder-template-name: coder
+          coder-template-preset: ${{ steps.determine-context.outputs.template_preset }}
+          coder-task-name-prefix: doc-check
+          coder-task-prompt: ${{ steps.extract-context.outputs.task_prompt }}
+          github-user-id: ${{ steps.determine-context.outputs.github_user_id }}
+          github-token: ${{ github.token }}
+          github-issue-url: ${{ steps.determine-context.outputs.pr_url }}
+          comment-on-issue: true
+
+      - name: Write outputs
+        env:
+          TASK_CREATED: ${{ steps.create_task.outputs.task-created }}
+          TASK_NAME: ${{ steps.create_task.outputs.task-name }}
+          TASK_URL: ${{ steps.create_task.outputs.task-url }}
+          PR_URL: ${{ steps.determine-context.outputs.pr_url }}
+        run: |
+          {
+            echo "## Documentation Check Task"
+            echo ""
+            echo "**PR:** ${PR_URL}"
+            echo "**Task created:** ${TASK_CREATED}"
+            echo "**Task name:** ${TASK_NAME}"
+            echo "**Task URL:** ${TASK_URL}"
+            echo ""
+            echo "The Coder task is analyzing the PR changes and will comment with documentation recommendations."
+          } >> "${GITHUB_STEP_SUMMARY}"

.github/workflows/docker-base.yaml

@@ -38,15 +38,17 @@ jobs:
     if: github.repository_owner == 'coder'
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Docker login
-        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3.4.0
+        uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # v3.6.0
         with:
           registry: ghcr.io
           username: ${{ github.actor }}
@@ -60,7 +62,7 @@ jobs:
 
       # This uses OIDC authentication, so no auth variables are required.
       - name: Build base Docker image via depot.dev
-        uses: depot/build-push-action@2583627a84956d07561420dcc1d0eb1f2af3fac0 # v1.15.0
+        uses: depot/build-push-action@9785b135c3c76c33db102e45be96a25ab55cd507 # v1.16.2
         with:
           project: wl5hnrrkns
           context: base-build-context

.github/workflows/docs-ci.yaml

@@ -23,12 +23,14 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Setup Node
         uses: ./.github/actions/setup-node
 
-      - uses: tj-actions/changed-files@666c9d29007687c52e3c7aa2aac6c0ffcadeadc3 # v45.0.7
+      - uses: tj-actions/changed-files@abdd2f68ea150cee8f236d4a9fb4e0f2491abf1b # v45.0.7
         id: changed-files
         with:
           files: |
@@ -39,10 +41,16 @@ jobs:
       - name: lint
         if: steps.changed-files.outputs.any_changed == 'true'
         run: |
-          pnpm exec markdownlint-cli2 ${{ steps.changed-files.outputs.all_changed_files }}
+          # shellcheck disable=SC2086
+          pnpm exec markdownlint-cli2 $ALL_CHANGED_FILES
+        env:
+          ALL_CHANGED_FILES: ${{ steps.changed-files.outputs.all_changed_files }}
 
       - name: fmt
         if: steps.changed-files.outputs.any_changed == 'true'
         run: |
           # markdown-table-formatter requires a space separated list of files
-          echo ${{ steps.changed-files.outputs.all_changed_files }} | tr ',' '\n' | pnpm exec markdown-table-formatter --check
+          # shellcheck disable=SC2086
+          echo $ALL_CHANGED_FILES | tr ',' '\n' | pnpm exec markdown-table-formatter --check
+        env:
+          ALL_CHANGED_FILES: ${{ steps.changed-files.outputs.all_changed_files }}

.github/workflows/dogfood.yaml

@@ -18,8 +18,7 @@ on:
   workflow_dispatch:
 
 permissions:
-  # Necessary for GCP authentication (https://github.com/google-github-actions/setup-gcloud#usage)
-  id-token: write
+  contents: read
 
 jobs:
   build_image:
@@ -27,19 +26,21 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-4' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Setup Nix
-        uses: nixbuild/nix-quick-install-action@63ca48f939ee3b8d835f4126562537df0fee5b91 # v32
+        uses: nixbuild/nix-quick-install-action@2c9db80fb984ceb1bcaa77cdda3fdf8cfba92035 # v34
         with:
           # Pinning to 2.28 here, as Nix gets a "error: [json.exception.type_error.302] type must be array, but is string"
           # on version 2.29 and above.
-          nix_version: "2.28.4"
+          nix_version: "2.28.5"
 
       - uses: nix-community/cache-nix-action@135667ec418502fa5a3598af6fb9eb733888ce6a # v6.1.3
         with:
@@ -62,15 +63,16 @@ jobs:
 
       - name: Get branch name
         id: branch-name
-        uses: tj-actions/branch-names@dde14ac574a8b9b1cedc59a1cf312788af43d8d8 # v8.2.1
+        uses: tj-actions/branch-names@5250492686b253f06fa55861556d1027b067aeb5 # v9.0.2
 
       - name: "Branch name to Docker tag name"
         id: docker-tag-name
         run: |
-          tag=${{ steps.branch-name.outputs.current_branch }}
           # Replace / with --, e.g. user/feature => user--feature.
-          tag=${tag//\//--}
-          echo "tag=${tag}" >> $GITHUB_OUTPUT
+          tag=${BRANCH_NAME//\//--}
+          echo "tag=${tag}" >> "$GITHUB_OUTPUT"
+        env:
+          BRANCH_NAME: ${{ steps.branch-name.outputs.current_branch }}
 
       - name: Set up Depot CLI
         uses: depot/setup-action@b0b1ea4f69e92ebf5dea3f8713a1b0c37b2126a5 # v1.6.0
@@ -80,13 +82,13 @@ jobs:
 
       - name: Login to DockerHub
         if: github.ref == 'refs/heads/main'
-        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3.4.0
+        uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # v3.6.0
         with:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_PASSWORD }}
 
       - name: Build and push Non-Nix image
-        uses: depot/build-push-action@2583627a84956d07561420dcc1d0eb1f2af3fac0 # v1.15.0
+        uses: depot/build-push-action@9785b135c3c76c33db102e45be96a25ab55cd507 # v1.16.2
         with:
           project: b4q6ltmpzh
           token: ${{ secrets.DEPOT_TOKEN }}
@@ -107,29 +109,36 @@ jobs:
 
           CURRENT_SYSTEM=$(nix eval --impure --raw --expr 'builtins.currentSystem')
 
-          docker image tag codercom/oss-dogfood-nix:latest-$CURRENT_SYSTEM codercom/oss-dogfood-nix:${{ steps.docker-tag-name.outputs.tag }}
-          docker image push codercom/oss-dogfood-nix:${{ steps.docker-tag-name.outputs.tag }}
+          docker image tag "codercom/oss-dogfood-nix:latest-$CURRENT_SYSTEM" "codercom/oss-dogfood-nix:${DOCKER_TAG}"
+          docker image push "codercom/oss-dogfood-nix:${DOCKER_TAG}"
 
-          docker image tag codercom/oss-dogfood-nix:latest-$CURRENT_SYSTEM codercom/oss-dogfood-nix:latest
-          docker image push codercom/oss-dogfood-nix:latest
+          docker image tag "codercom/oss-dogfood-nix:latest-$CURRENT_SYSTEM" "codercom/oss-dogfood-nix:latest"
+          docker image push "codercom/oss-dogfood-nix:latest"
+        env:
+          DOCKER_TAG: ${{ steps.docker-tag-name.outputs.tag }}
 
   deploy_template:
     needs: build_image
     runs-on: ubuntu-latest
+    permissions:
+      # Necessary for GCP authentication (https://github.com/google-github-actions/setup-gcloud#usage)
+      id-token: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Setup Terraform
         uses: ./.github/actions/setup-tf
 
       - name: Authenticate to Google Cloud
-        uses: google-github-actions/auth@ba79af03959ebeac9769e648f473a284504d9193 # v2.1.10
+        uses: google-github-actions/auth@7c6bc770dae815cd3e89ee6cdf493a5fab2cc093 # v3.0.0
         with:
           workload_identity_provider: ${{ vars.GCP_WORKLOAD_ID_PROVIDER }}
           service_account: ${{ vars.GCP_SERVICE_ACCOUNT }}
@@ -152,12 +161,12 @@ jobs:
       - name: Get short commit SHA
         if: github.ref == 'refs/heads/main'
         id: vars
-        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> "$GITHUB_OUTPUT"
 
       - name: Get latest commit title
         if: github.ref == 'refs/heads/main'
         id: message
-        run: echo "pr_title=$(git log --format=%s -n 1 ${{ github.sha }})" >> $GITHUB_OUTPUT
+        run: echo "pr_title=$(git log --format=%s -n 1 ${{ github.sha }})" >> "$GITHUB_OUTPUT"
 
       - name: "Push template"
         if: github.ref == 'refs/heads/main'
@@ -169,6 +178,7 @@ jobs:
           CODER_URL: https://dev.coder.com
           CODER_SESSION_TOKEN: ${{ secrets.CODER_SESSION_TOKEN }}
           # Template source & details
+          TF_VAR_CODER_DOGFOOD_ANTHROPIC_API_KEY: ${{ secrets.CODER_DOGFOOD_ANTHROPIC_API_KEY }}
           TF_VAR_CODER_TEMPLATE_NAME: ${{ secrets.CODER_TEMPLATE_NAME }}
           TF_VAR_CODER_TEMPLATE_VERSION: ${{ steps.vars.outputs.sha_short }}
           TF_VAR_CODER_TEMPLATE_DIR: ./coder

.github/workflows/nightly-gauntlet.yaml

@@ -0,0 +1,204 @@
+# The nightly-gauntlet runs tests that are either too flaky or too slow to block
+# every PR.
+name: nightly-gauntlet
+on:
+  schedule:
+    # Every day at 4AM
+    - cron: "0 4 * * 1-5"
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  test-go-pg:
+    # make sure to adjust NUM_PARALLEL_PACKAGES and NUM_PARALLEL_TESTS below
+    # when changing runner sizes
+    runs-on: ${{ matrix.os == 'macos-latest' && github.repository_owner == 'coder' && 'depot-macos-latest' || matrix.os == 'windows-2022' && github.repository_owner == 'coder' && 'depot-windows-2022-16' || matrix.os }}
+    # This timeout must be greater than the timeout set by `go test` in
+    # `make test-postgres` to ensure we receive a trace of running
+    # goroutines. Setting this to the timeout +5m should work quite well
+    # even if some of the preceding steps are slow.
+    timeout-minutes: 25
+    strategy:
+      matrix:
+        os:
+          - macos-latest
+          - windows-2022
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
+        with:
+          egress-policy: audit
+
+      # macOS indexes all new files in the background. Our Postgres tests
+      # create and destroy thousands of databases on disk, and Spotlight
+      # tries to index all of them, seriously slowing down the tests.
+      - name: Disable Spotlight Indexing
+        if: runner.os == 'macOS'
+        run: |
+          enabled=$(sudo mdutil -a -s | { grep -Fc "Indexing enabled" || true; })
+          if [ "$enabled" -eq 0 ]; then
+            echo "Spotlight indexing is already disabled"
+            exit 0
+          fi
+          sudo mdutil -a -i off
+          sudo mdutil -X /
+          sudo launchctl bootout system /System/Library/LaunchDaemons/com.apple.metadata.mds.plist
+
+      # Set up RAM disks to speed up the rest of the job. This action is in
+      # a separate repository to allow its use before actions/checkout.
+      - name: Setup RAM Disks
+        if: runner.os == 'Windows'
+        uses: coder/setup-ramdisk-action@e1100847ab2d7bcd9d14bcda8f2d1b0f07b36f1b # v0.1.0
+
+      - name: Checkout
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          fetch-depth: 1
+          persist-credentials: false
+
+      - name: Setup Go
+        uses: ./.github/actions/setup-go
+        with:
+          # Runners have Go baked-in and Go will automatically
+          # download the toolchain configured in go.mod, so we don't
+          # need to reinstall it. It's faster on Windows runners.
+          use-preinstalled-go: ${{ runner.os == 'Windows' }}
+
+      - name: Setup Terraform
+        uses: ./.github/actions/setup-tf
+
+      - name: Setup Embedded Postgres Cache Paths
+        id: embedded-pg-cache
+        uses: ./.github/actions/setup-embedded-pg-cache-paths
+
+      - name: Download Embedded Postgres Cache
+        id: download-embedded-pg-cache
+        uses: ./.github/actions/embedded-pg-cache/download
+        with:
+          key-prefix: embedded-pg-${{ runner.os }}-${{ runner.arch }}
+          cache-path: ${{ steps.embedded-pg-cache.outputs.cached-dirs }}
+
+      - name: Test with PostgreSQL Database
+        env:
+          POSTGRES_VERSION: "13"
+          TS_DEBUG_DISCO: "true"
+          LC_CTYPE: "en_US.UTF-8"
+          LC_ALL: "en_US.UTF-8"
+        shell: bash
+        run: |
+          set -o errexit
+          set -o pipefail
+
+          if [ "${{ runner.os }}" == "Windows" ]; then
+            # Create a temp dir on the R: ramdisk drive for Windows. The default
+            # C: drive is extremely slow: https://github.com/actions/runner-images/issues/8755
+            mkdir -p "R:/temp/embedded-pg"
+            go run scripts/embedded-pg/main.go -path "R:/temp/embedded-pg" -cache "${EMBEDDED_PG_CACHE_DIR}"
+          elif [ "${{ runner.os }}" == "macOS" ]; then
+            # Postgres runs faster on a ramdisk on macOS too
+            mkdir -p /tmp/tmpfs
+            sudo mount_tmpfs -o noowners -s 8g /tmp/tmpfs
+            go run scripts/embedded-pg/main.go -path /tmp/tmpfs/embedded-pg -cache "${EMBEDDED_PG_CACHE_DIR}"
+          elif [ "${{ runner.os }}" == "Linux" ]; then
+            make test-postgres-docker
+          fi
+
+          # if macOS, install google-chrome for scaletests
+          # As another concern, should we really have this kind of external dependency
+          # requirement on standard CI?
+          if [ "${{ matrix.os }}" == "macos-latest" ]; then
+            brew install google-chrome
+          fi
+
+          # macOS will output "The default interactive shell is now zsh"
+          # intermittently in CI...
+          if [ "${{ matrix.os }}" == "macos-latest" ]; then
+            touch ~/.bash_profile && echo "export BASH_SILENCE_DEPRECATION_WARNING=1" >> ~/.bash_profile
+          fi
+
+          if [ "${{ runner.os }}" == "Windows" ]; then
+            # Our Windows runners have 16 cores.
+            # On Windows Postgres chokes up when we have 16x16=256 tests
+            # running in parallel, and dbtestutil.NewDB starts to take more than
+            # 10s to complete sometimes causing test timeouts. With 16x8=128 tests
+            # Postgres tends not to choke.
+            NUM_PARALLEL_PACKAGES=8
+            NUM_PARALLEL_TESTS=16
+          elif [ "${{ runner.os }}" == "macOS" ]; then
+            # Our macOS runners have 8 cores. We set NUM_PARALLEL_TESTS to 16
+            # because the tests complete faster and Postgres doesn't choke. It seems
+            # that macOS's tmpfs is faster than the one on Windows.
+            NUM_PARALLEL_PACKAGES=8
+            NUM_PARALLEL_TESTS=16
+          elif [ "${{ runner.os }}" == "Linux" ]; then
+            # Our Linux runners have 8 cores.
+            NUM_PARALLEL_PACKAGES=8
+            NUM_PARALLEL_TESTS=8
+          fi
+
+          # run tests without cache
+          TESTCOUNT="-count=1"
+
+          DB=ci gotestsum \
+            --format standard-quiet --packages "./..." \
+            -- -timeout=20m -v -p "$NUM_PARALLEL_PACKAGES" -parallel="$NUM_PARALLEL_TESTS" "$TESTCOUNT"
+
+      - name: Upload Embedded Postgres Cache
+        uses: ./.github/actions/embedded-pg-cache/upload
+        # We only use the embedded Postgres cache on macOS and Windows runners.
+        if: runner.OS == 'macOS' || runner.OS == 'Windows'
+        with:
+          cache-key: ${{ steps.download-embedded-pg-cache.outputs.cache-key }}
+          cache-path: "${{ steps.embedded-pg-cache.outputs.embedded-pg-cache }}"
+
+      - name: Upload test stats to Datadog
+        timeout-minutes: 1
+        continue-on-error: true
+        uses: ./.github/actions/upload-datadog
+        if: success() || failure()
+        with:
+          api-key: ${{ secrets.DATADOG_API_KEY }}
+
+  notify-slack-on-failure:
+    needs:
+      - test-go-pg
+    runs-on: ubuntu-latest
+    if: failure() && github.ref == 'refs/heads/main'
+
+    steps:
+      - name: Send Slack notification
+        run: |
+          ESCAPED_PROMPT=$(printf "%s" "<@U09LQ75AHKR> $BLINK_CI_FAILURE_PROMPT" | jq -Rsa .)
+          curl -X POST -H 'Content-type: application/json' \
+          --data '{
+            "blocks": [
+              {
+                "type": "header",
+                "text": {
+                  "type": "plain_text",
+                  "text": "❌ Nightly gauntlet failed",
+                  "emoji": true
+                }
+              },
+              {
+                "type": "section",
+                "text": {
+                  "type": "mrkdwn",
+                  "text": "*View failure:* <'"${RUN_URL}"'|Click here>"
+                }
+              },
+              {
+                "type": "section",
+                "text": {
+                  "type": "mrkdwn",
+                  "text": '"$ESCAPED_PROMPT"'
+                }
+              }
+            ]
+          }' "${SLACK_WEBHOOK}"
+        env:
+          SLACK_WEBHOOK: ${{ secrets.CI_FAILURE_SLACK_WEBHOOK }}
+          RUN_URL: "${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+          BLINK_CI_FAILURE_PROMPT: ${{ vars.BLINK_CI_FAILURE_PROMPT }}

.github/workflows/pr-auto-assign.yaml

@@ -3,6 +3,7 @@
 name: PR Auto Assign
 
 on:
+  # zizmor: ignore[dangerous-triggers] We explicitly want to run on pull_request_target.
   pull_request_target:
     types: [opened]
 
@@ -14,7 +15,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 

.github/workflows/pr-cleanup.yaml

@@ -19,7 +19,7 @@ jobs:
       packages: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
@@ -27,10 +27,12 @@ jobs:
         id: pr_number
         run: |
           if [ -n "${{ github.event.pull_request.number }}" ]; then
-            echo "PR_NUMBER=${{ github.event.pull_request.number }}" >> $GITHUB_OUTPUT
+            echo "PR_NUMBER=${{ github.event.pull_request.number }}" >> "$GITHUB_OUTPUT"
           else
-            echo "PR_NUMBER=${{ github.event.inputs.pr_number }}" >> $GITHUB_OUTPUT
+            echo "PR_NUMBER=${PR_NUMBER}" >> "$GITHUB_OUTPUT"
           fi
+        env:
+          PR_NUMBER: ${{ github.event.inputs.pr_number }}
 
       - name: Delete image
         continue-on-error: true
@@ -51,17 +53,21 @@ jobs:
       - name: Delete helm release
         run: |
           set -euo pipefail
-          helm delete --namespace "pr${{ steps.pr_number.outputs.PR_NUMBER }}" "pr${{ steps.pr_number.outputs.PR_NUMBER }}" || echo "helm release not found"
+          helm delete --namespace "pr${PR_NUMBER}" "pr${PR_NUMBER}" || echo "helm release not found"
+        env:
+          PR_NUMBER: ${{ steps.pr_number.outputs.PR_NUMBER }}
 
       - name: "Remove PR namespace"
         run: |
-          kubectl delete namespace "pr${{ steps.pr_number.outputs.PR_NUMBER }}" || echo "namespace not found"
+          kubectl delete namespace "pr${PR_NUMBER}" || echo "namespace not found"
+        env:
+          PR_NUMBER: ${{ steps.pr_number.outputs.PR_NUMBER }}
 
       - name: "Remove DNS records"
         run: |
           set -euo pipefail
           # Get identifier for the record
-          record_id=$(curl -X GET "https://api.cloudflare.com/client/v4/zones/${{ secrets.PR_DEPLOYMENTS_ZONE_ID }}/dns_records?name=%2A.pr${{ steps.pr_number.outputs.PR_NUMBER }}.${{ secrets.PR_DEPLOYMENTS_DOMAIN }}" \
+          record_id=$(curl -X GET "https://api.cloudflare.com/client/v4/zones/${{ secrets.PR_DEPLOYMENTS_ZONE_ID }}/dns_records?name=%2A.pr${PR_NUMBER}.${{ secrets.PR_DEPLOYMENTS_DOMAIN }}" \
           -H "Authorization: Bearer ${{ secrets.PR_DEPLOYMENTS_CLOUDFLARE_API_TOKEN }}" \
           -H "Content-Type:application/json" | jq -r '.result[0].id') || echo "DNS record not found"
 
@@ -73,9 +79,13 @@ jobs:
             -H "Authorization: Bearer ${{ secrets.PR_DEPLOYMENTS_CLOUDFLARE_API_TOKEN }}" \
             -H "Content-Type:application/json" | jq -r '.success'
           ) || echo "DNS record not found"
+        env:
+          PR_NUMBER: ${{ steps.pr_number.outputs.PR_NUMBER }}
 
       - name: "Delete certificate"
         if: ${{ github.event.pull_request.merged == true }}
         run: |
           set -euxo pipefail
-          kubectl delete certificate "pr${{ steps.pr_number.outputs.PR_NUMBER }}-tls" -n pr-deployment-certs || echo "certificate not found"
+          kubectl delete certificate "pr${PR_NUMBER}-tls" -n pr-deployment-certs || echo "certificate not found"
+        env:
+          PR_NUMBER: ${{ steps.pr_number.outputs.PR_NUMBER }}

.github/workflows/pr-deploy.yaml

@@ -39,12 +39,14 @@ jobs:
       PR_OPEN: ${{ steps.check_pr.outputs.pr_open }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Check if PR is open
         id: check_pr
@@ -55,7 +57,7 @@ jobs:
             echo "PR doesn't exist or is closed."
             pr_open=false
           fi
-          echo "pr_open=$pr_open" >> $GITHUB_OUTPUT
+          echo "pr_open=$pr_open" >> "$GITHUB_OUTPUT"
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
@@ -74,14 +76,15 @@ jobs:
     runs-on: "ubuntu-latest"
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           fetch-depth: 0
+          persist-credentials: false
 
       - name: Get PR number, title, and branch name
         id: pr_info
@@ -90,9 +93,11 @@ jobs:
           PR_NUMBER=$(gh pr view --json number | jq -r '.number')
           PR_TITLE=$(gh pr view --json title | jq -r '.title')
           PR_URL=$(gh pr view --json url | jq -r '.url')
-          echo "PR_URL=$PR_URL" >> $GITHUB_OUTPUT
-          echo "PR_NUMBER=$PR_NUMBER" >> $GITHUB_OUTPUT
-          echo "PR_TITLE=$PR_TITLE" >> $GITHUB_OUTPUT
+          {
+            echo "PR_URL=$PR_URL"
+            echo "PR_NUMBER=$PR_NUMBER"
+            echo "PR_TITLE=$PR_TITLE"
+          } >> "$GITHUB_OUTPUT"
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
@@ -100,8 +105,8 @@ jobs:
         id: set_tags
         run: |
           set -euo pipefail
-          echo "CODER_BASE_IMAGE_TAG=$CODER_BASE_IMAGE_TAG" >> $GITHUB_OUTPUT
-          echo "CODER_IMAGE_TAG=$CODER_IMAGE_TAG" >> $GITHUB_OUTPUT
+          echo "CODER_BASE_IMAGE_TAG=$CODER_BASE_IMAGE_TAG" >> "$GITHUB_OUTPUT"
+          echo "CODER_IMAGE_TAG=$CODER_IMAGE_TAG" >> "$GITHUB_OUTPUT"
         env:
           CODER_BASE_IMAGE_TAG: ghcr.io/coder/coder-preview-base:pr${{ steps.pr_info.outputs.PR_NUMBER }}
           CODER_IMAGE_TAG: ghcr.io/coder/coder-preview:pr${{ steps.pr_info.outputs.PR_NUMBER }}
@@ -118,14 +123,16 @@ jobs:
         id: check_deployment
         run: |
           set -euo pipefail
-          if helm status "pr${{ steps.pr_info.outputs.PR_NUMBER }}" --namespace "pr${{ steps.pr_info.outputs.PR_NUMBER }}" > /dev/null 2>&1; then
+          if helm status "pr${PR_NUMBER}" --namespace "pr${PR_NUMBER}" > /dev/null 2>&1; then
             echo "Deployment already exists. Skipping deployment."
             NEW=false
           else
             echo "Deployment doesn't exist."
             NEW=true
           fi
-          echo "NEW=$NEW" >> $GITHUB_OUTPUT
+          echo "NEW=$NEW" >> "$GITHUB_OUTPUT"
+        env:
+          PR_NUMBER: ${{ steps.pr_info.outputs.PR_NUMBER }}
 
       - name: Check changed files
         uses: dorny/paths-filter@de90cc6fb38fc0963ad72b210f1f284cd68cea36 # v3.0.2
@@ -154,17 +161,20 @@ jobs:
       - name: Print number of changed files
         run: |
           set -euo pipefail
-          echo "Total number of changed files: ${{ steps.filter.outputs.all_count }}"
-          echo "Number of ignored files: ${{ steps.filter.outputs.ignored_count }}"
+          echo "Total number of changed files: ${ALL_COUNT}"
+          echo "Number of ignored files: ${IGNORED_COUNT}"
+        env:
+          ALL_COUNT: ${{ steps.filter.outputs.all_count }}
+          IGNORED_COUNT: ${{ steps.filter.outputs.ignored_count }}
 
       - name: Build conditionals
         id: build_conditionals
         run: |
           set -euo pipefail
           # build if the workflow is manually triggered and the deployment doesn't exist (first build or force rebuild)
-          echo "first_or_force_build=${{ (github.event_name == 'workflow_dispatch' && steps.check_deployment.outputs.NEW == 'true') || github.event.inputs.build == 'true' }}" >> $GITHUB_OUTPUT
+          echo "first_or_force_build=${{ (github.event_name == 'workflow_dispatch' && steps.check_deployment.outputs.NEW == 'true') || github.event.inputs.build == 'true' }}" >> "$GITHUB_OUTPUT"
           # build if the deployment already exist and there are changes in the files that we care about (automatic updates)
-          echo "automatic_rebuild=${{ steps.check_deployment.outputs.NEW == 'false' && steps.filter.outputs.all_count > steps.filter.outputs.ignored_count }}" >> $GITHUB_OUTPUT
+          echo "automatic_rebuild=${{ steps.check_deployment.outputs.NEW == 'false' && steps.filter.outputs.all_count > steps.filter.outputs.ignored_count }}" >> "$GITHUB_OUTPUT"
 
   comment-pr:
     needs: get_info
@@ -174,12 +184,12 @@ jobs:
       pull-requests: write # needed for commenting on PRs
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Find Comment
-        uses: peter-evans/find-comment@3eae4d37986fb5a8592848f6a574fdf654e61f9e # v3.1.0
+        uses: peter-evans/find-comment@b30e6a3c0ed37e7c023ccd3f1db5c6c0b0c23aad # v4.0.0
         id: fc
         with:
           issue-number: ${{ needs.get_info.outputs.PR_NUMBER }}
@@ -189,7 +199,7 @@ jobs:
 
       - name: Comment on PR
         id: comment_id
-        uses: peter-evans/create-or-update-comment@71345be0265236311c031f5c7866368bd1eff043 # v4.0.0
+        uses: peter-evans/create-or-update-comment@e8674b075228eee787fea43ef493e45ece1004c9 # v5.0.0
         with:
           comment-id: ${{ steps.fc.outputs.comment-id }}
           issue-number: ${{ needs.get_info.outputs.PR_NUMBER }}
@@ -218,14 +228,15 @@ jobs:
       CODER_IMAGE_TAG: ${{ needs.get_info.outputs.CODER_IMAGE_TAG }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           fetch-depth: 0
+          persist-credentials: false
 
       - name: Setup Node
         uses: ./.github/actions/setup-node
@@ -237,7 +248,7 @@ jobs:
         uses: ./.github/actions/setup-sqlc
 
       - name: GHCR Login
-        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3.4.0
+        uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # v3.6.0
         with:
           registry: ghcr.io
           username: ${{ github.actor }}
@@ -250,12 +261,13 @@ jobs:
           make gen/mark-fresh
           export DOCKER_IMAGE_NO_PREREQUISITES=true
           version="$(./scripts/version.sh)"
-          export CODER_IMAGE_BUILD_BASE_TAG="$(CODER_IMAGE_BASE=coder-base ./scripts/image_tag.sh --version "$version")"
+          CODER_IMAGE_BUILD_BASE_TAG="$(CODER_IMAGE_BASE=coder-base ./scripts/image_tag.sh --version "$version")"
+          export CODER_IMAGE_BUILD_BASE_TAG
           make -j build/coder_linux_amd64
           ./scripts/build_docker.sh \
             --arch amd64 \
-            --target ${{ env.CODER_IMAGE_TAG }} \
-            --version $version \
+            --target "${CODER_IMAGE_TAG}" \
+            --version "$version" \
             --push \
             build/coder_linux_amd64
 
@@ -276,7 +288,7 @@ jobs:
       PR_HOSTNAME: "pr${{ needs.get_info.outputs.PR_NUMBER }}.${{ secrets.PR_DEPLOYMENTS_DOMAIN }}"
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
@@ -293,13 +305,13 @@ jobs:
           set -euo pipefail
           foundTag=$(
             gh api /orgs/coder/packages/container/coder-preview/versions |
-            jq -r --arg tag "pr${{ env.PR_NUMBER }}" '.[] |
+            jq -r --arg tag "pr${PR_NUMBER}" '.[] |
             select(.metadata.container.tags == [$tag]) |
             .metadata.container.tags[0]'
           )
           if [ -z "$foundTag" ]; then
             echo "Image not found"
-            echo "${{ env.CODER_IMAGE_TAG }} not found in ghcr.io/coder/coder-preview"
+            echo "${CODER_IMAGE_TAG} not found in ghcr.io/coder/coder-preview"
             exit 1
           else
             echo "Image found"
@@ -314,40 +326,42 @@ jobs:
           curl -X POST "https://api.cloudflare.com/client/v4/zones/${{ secrets.PR_DEPLOYMENTS_ZONE_ID }}/dns_records" \
             -H "Authorization: Bearer ${{ secrets.PR_DEPLOYMENTS_CLOUDFLARE_API_TOKEN }}" \
             -H "Content-Type:application/json" \
-            --data '{"type":"CNAME","name":"*.${{ env.PR_HOSTNAME }}","content":"${{ env.PR_HOSTNAME }}","ttl":1,"proxied":false}'
+            --data '{"type":"CNAME","name":"*.'"${PR_HOSTNAME}"'","content":"'"${PR_HOSTNAME}"'","ttl":1,"proxied":false}'
 
       - name: Create PR namespace
         if: needs.get_info.outputs.NEW == 'true' || github.event.inputs.deploy == 'true'
         run: |
           set -euo pipefail
           # try to delete the namespace, but don't fail if it doesn't exist
-          kubectl delete namespace "pr${{ env.PR_NUMBER }}" || true
-          kubectl create namespace "pr${{ env.PR_NUMBER }}"
+          kubectl delete namespace "pr${PR_NUMBER}" || true
+          kubectl create namespace "pr${PR_NUMBER}"
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Check and Create Certificate
         if: needs.get_info.outputs.NEW == 'true' || github.event.inputs.deploy == 'true'
         run: |
           # Using kubectl to check if a Certificate resource already exists
           # we are doing this to avoid letsenrypt rate limits
-          if ! kubectl get certificate pr${{ env.PR_NUMBER }}-tls -n pr-deployment-certs > /dev/null 2>&1; then
+          if ! kubectl get certificate "pr${PR_NUMBER}-tls" -n pr-deployment-certs > /dev/null 2>&1; then
             echo "Certificate doesn't exist. Creating a new one."
             envsubst < ./.github/pr-deployments/certificate.yaml | kubectl apply -f -
           else
             echo "Certificate exists. Skipping certificate creation."
           fi
-          echo "Copy certificate from pr-deployment-certs to pr${{ env.PR_NUMBER }} namespace"
-          until kubectl get secret pr${{ env.PR_NUMBER }}-tls -n pr-deployment-certs &> /dev/null
+          echo "Copy certificate from pr-deployment-certs to pr${PR_NUMBER} namespace"
+          until kubectl get secret "pr${PR_NUMBER}-tls" -n pr-deployment-certs &> /dev/null
           do
-            echo "Waiting for secret pr${{ env.PR_NUMBER }}-tls to be created..."
+            echo "Waiting for secret pr${PR_NUMBER}-tls to be created..."
             sleep 5
           done
           (
-            kubectl get secret pr${{ env.PR_NUMBER }}-tls -n pr-deployment-certs -o json |
+            kubectl get secret "pr${PR_NUMBER}-tls" -n pr-deployment-certs -o json |
             jq 'del(.metadata.namespace,.metadata.creationTimestamp,.metadata.resourceVersion,.metadata.selfLink,.metadata.uid,.metadata.managedFields)' |
-            kubectl -n pr${{ env.PR_NUMBER }} apply -f -
+            kubectl -n "pr${PR_NUMBER}" apply -f -
           )
 
       - name: Set up PostgreSQL database
@@ -355,13 +369,14 @@ jobs:
         run: |
           helm repo add bitnami https://charts.bitnami.com/bitnami
           helm install coder-db bitnami/postgresql \
-            --namespace pr${{ env.PR_NUMBER }} \
+            --namespace "pr${PR_NUMBER}" \
+            --set image.repository=bitnamilegacy/postgresql \
             --set auth.username=coder \
             --set auth.password=coder \
             --set auth.database=coder \
             --set persistence.size=10Gi
-          kubectl create secret generic coder-db-url -n pr${{ env.PR_NUMBER }} \
-            --from-literal=url="postgres://coder:coder@coder-db-postgresql.pr${{ env.PR_NUMBER }}.svc.cluster.local:5432/coder?sslmode=disable"
+          kubectl create secret generic coder-db-url -n "pr${PR_NUMBER}" \
+            --from-literal=url="postgres://coder:coder@coder-db-postgresql.pr${PR_NUMBER}.svc.cluster.local:5432/coder?sslmode=disable"
 
       - name: Create a service account, role, and rolebinding for the PR namespace
         if: needs.get_info.outputs.NEW == 'true' || github.event.inputs.deploy == 'true'
@@ -383,8 +398,8 @@ jobs:
         run: |
           set -euo pipefail
           helm dependency update --skip-refresh ./helm/coder
-          helm upgrade --install "pr${{ env.PR_NUMBER }}" ./helm/coder \
-          --namespace "pr${{ env.PR_NUMBER }}" \
+          helm upgrade --install "pr${PR_NUMBER}" ./helm/coder \
+          --namespace "pr${PR_NUMBER}" \
           --values ./pr-deploy-values.yaml \
           --force
 
@@ -393,8 +408,8 @@ jobs:
         run: |
           helm repo add coder-logstream-kube https://helm.coder.com/logstream-kube
           helm upgrade --install coder-logstream-kube coder-logstream-kube/coder-logstream-kube \
-            --namespace "pr${{ env.PR_NUMBER }}" \
-            --set url="https://${{ env.PR_HOSTNAME }}"
+            --namespace "pr${PR_NUMBER}" \
+            --set url="https://${PR_HOSTNAME}"
 
       - name: Get Coder binary
         if: needs.get_info.outputs.NEW == 'true' || github.event.inputs.deploy == 'true'
@@ -402,16 +417,16 @@ jobs:
           set -euo pipefail
 
           DEST="${HOME}/coder"
-          URL="https://${{ env.PR_HOSTNAME }}/bin/coder-linux-amd64"
+          URL="https://${PR_HOSTNAME}/bin/coder-linux-amd64"
 
-          mkdir -p "$(dirname ${DEST})"
+          mkdir -p "$(dirname "$DEST")"
 
           COUNT=0
-          until $(curl --output /dev/null --silent --head --fail "$URL"); do
+          until curl --output /dev/null --silent --head --fail "$URL"; do
               printf '.'
               sleep 5
               COUNT=$((COUNT+1))
-              if [ $COUNT -ge 60 ]; then
+              if [ "$COUNT" -ge 60 ]; then
                 echo "Timed out waiting for URL to be available"
                 exit 1
               fi
@@ -435,24 +450,24 @@ jobs:
 
           # add mask so that the password is not printed to the logs
           echo "::add-mask::$password"
-          echo "password=$password" >> $GITHUB_OUTPUT
+          echo "password=$password" >> "$GITHUB_OUTPUT"
 
           coder login \
-            --first-user-username pr${{ env.PR_NUMBER }}-admin \
-            --first-user-email pr${{ env.PR_NUMBER }}@coder.com \
-            --first-user-password $password \
+            --first-user-username "pr${PR_NUMBER}-admin" \
+            --first-user-email "pr${PR_NUMBER}@coder.com" \
+            --first-user-password "$password" \
             --first-user-trial=false \
             --use-token-as-session \
-            https://${{ env.PR_HOSTNAME }}
+            "https://${PR_HOSTNAME}"
 
           # Create a user for the github.actor
           # TODO: update once https://github.com/coder/coder/issues/15466 is resolved
           # coder users create \
-          #   --username ${{ github.actor }} \
+          #   --username ${GITHUB_ACTOR} \
           #   --login-type github
 
           # promote the user to admin role
-          # coder org members edit-role ${{ github.actor }} organization-admin
+          # coder org members edit-role ${GITHUB_ACTOR} organization-admin
           # TODO: update once https://github.com/coder/internal/issues/207 is resolved
 
       - name: Send Slack notification
@@ -461,20 +476,22 @@ jobs:
           curl -s -o /dev/null -X POST -H 'Content-type: application/json' \
             -d \
             '{
-              "pr_number": "'"${{ env.PR_NUMBER }}"'",
-              "pr_url": "'"${{ env.PR_URL }}"'",
-              "pr_title": "'"${{ env.PR_TITLE }}"'",
-              "pr_access_url": "'"https://${{ env.PR_HOSTNAME }}"'",
-              "pr_username": "'"pr${{ env.PR_NUMBER }}-admin"'",
-              "pr_email": "'"pr${{ env.PR_NUMBER }}@coder.com"'",
-              "pr_password": "'"${{ steps.setup_deployment.outputs.password }}"'",
-              "pr_actor": "'"${{ github.actor }}"'"
+              "pr_number": "'"${PR_NUMBER}"'",
+              "pr_url": "'"${PR_URL}"'",
+              "pr_title": "'"${PR_TITLE}"'",
+              "pr_access_url": "'"https://${PR_HOSTNAME}"'",
+              "pr_username": "'"pr${PR_NUMBER}-admin"'",
+              "pr_email": "'"pr${PR_NUMBER}@coder.com"'",
+              "pr_password": "'"${PASSWORD}"'",
+              "pr_actor": "'"${GITHUB_ACTOR}"'"
             }' \
             ${{ secrets.PR_DEPLOYMENTS_SLACK_WEBHOOK }}
           echo "Slack notification sent"
+        env:
+          PASSWORD: ${{ steps.setup_deployment.outputs.password }}
 
       - name: Find Comment
-        uses: peter-evans/find-comment@3eae4d37986fb5a8592848f6a574fdf654e61f9e # v3.1.0
+        uses: peter-evans/find-comment@b30e6a3c0ed37e7c023ccd3f1db5c6c0b0c23aad # v4.0.0
         id: fc
         with:
           issue-number: ${{ env.PR_NUMBER }}
@@ -483,7 +500,7 @@ jobs:
           direction: last
 
       - name: Comment on PR
-        uses: peter-evans/create-or-update-comment@71345be0265236311c031f5c7866368bd1eff043 # v4.0.0
+        uses: peter-evans/create-or-update-comment@e8674b075228eee787fea43ef493e45ece1004c9 # v5.0.0
         env:
           STATUS: ${{ needs.get_info.outputs.NEW == 'true' && 'Created' || 'Updated' }}
         with:
@@ -504,7 +521,7 @@ jobs:
         run: |
           set -euo pipefail
           cd .github/pr-deployments/template
-          coder templates push -y --variable namespace=pr${{ env.PR_NUMBER }} kubernetes
+          coder templates push -y --variable "namespace=pr${PR_NUMBER}" kubernetes
 
           # Create workspace
           coder create --template="kubernetes" kube --parameter cpu=2 --parameter memory=4 --parameter home_disk_size=2 -y

.github/workflows/release-validation.yaml

@@ -14,7 +14,7 @@ jobs:
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 

.github/workflows/release.yaml

@@ -32,15 +32,43 @@ env:
   CODER_RELEASE_NOTES: ${{ inputs.release_notes }}
 
 jobs:
+  # Only allow maintainers/admins to release.
+  check-perms:
+    runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
+    steps:
+      - name: Allow only maintainers/admins
+        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
+        with:
+          github-token: ${{ secrets.GITHUB_TOKEN }}
+          script: |
+            const {data} = await github.rest.repos.getCollaboratorPermissionLevel({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              username: context.actor
+            });
+            const role = data.role_name || data.user?.role_name || data.permission;
+            const perms = data.user?.permissions || {};
+            core.info(`Actor ${context.actor} permission=${data.permission}, role_name=${role}`);
+
+            const allowed =
+              role === 'admin' ||
+              role === 'maintain' ||
+              perms.admin === true ||
+              perms.maintain === true;
+
+            if (!allowed) core.setFailed('Denied: requires maintain or admin');
+
   # build-dylib is a separate job to build the dylib on macOS.
   build-dylib:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-macos-latest' || 'macos-latest' }}
+    needs: check-perms
     steps:
       # Harden Runner doesn't work on macOS.
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           fetch-depth: 0
+          persist-credentials: false
 
       # If the event that triggered the build was an annotated tag (which our
       # tags are supposed to be), actions/checkout has a bug where the tag in
@@ -53,9 +81,11 @@ jobs:
       - name: Setup build tools
         run: |
           brew install bash gnu-getopt make
-          echo "$(brew --prefix bash)/bin" >> $GITHUB_PATH
-          echo "$(brew --prefix gnu-getopt)/bin" >> $GITHUB_PATH
-          echo "$(brew --prefix make)/libexec/gnubin" >> $GITHUB_PATH
+          {
+            echo "$(brew --prefix bash)/bin"
+            echo "$(brew --prefix gnu-getopt)/bin"
+            echo "$(brew --prefix make)/libexec/gnubin"
+          } >> "$GITHUB_PATH"
 
       - name: Switch XCode Version
         uses: maxim-lobanov/setup-xcode@60606e260d2fc5762a71e64e74b2174e8ea3c8bd # v1.6.0
@@ -101,7 +131,7 @@ jobs:
           AC_CERTIFICATE_PASSWORD_FILE: /tmp/apple_cert_password.txt
 
       - name: Upload build artifacts
-        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: dylibs
           path: |
@@ -114,7 +144,7 @@ jobs:
 
   release:
     name: Build and publish
-    needs: build-dylib
+    needs: [build-dylib, check-perms]
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
     permissions:
       # Required to publish a release
@@ -134,14 +164,15 @@ jobs:
       version: ${{ steps.version.outputs.version }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           fetch-depth: 0
+          persist-credentials: false
 
       # If the event that triggered the build was an annotated tag (which our
       # tags are supposed to be), actions/checkout has a bug where the tag in
@@ -156,9 +187,9 @@ jobs:
         run: |
           set -euo pipefail
           version="$(./scripts/version.sh)"
-          echo "version=$version" >> $GITHUB_OUTPUT
+          echo "version=$version" >> "$GITHUB_OUTPUT"
           # Speed up future version.sh calls.
-          echo "CODER_FORCE_VERSION=$version" >> $GITHUB_ENV
+          echo "CODER_FORCE_VERSION=$version" >> "$GITHUB_ENV"
           echo "$version"
 
       # Verify that all expectations for a release are met.
@@ -200,7 +231,7 @@ jobs:
 
           release_notes_file="$(mktemp -t release_notes.XXXXXX)"
           echo "$CODER_RELEASE_NOTES" > "$release_notes_file"
-          echo CODER_RELEASE_NOTES_FILE="$release_notes_file" >> $GITHUB_ENV
+          echo CODER_RELEASE_NOTES_FILE="$release_notes_file" >> "$GITHUB_ENV"
 
       - name: Show release notes
         run: |
@@ -208,7 +239,7 @@ jobs:
           cat "$CODER_RELEASE_NOTES_FILE"
 
       - name: Docker Login
-        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3.4.0
+        uses: docker/login-action@5e57cd118135c172c3672efd75eb46360885c0ef # v3.6.0
         with:
           registry: ghcr.io
           username: ${{ github.actor }}
@@ -222,7 +253,7 @@ jobs:
 
       # Necessary for signing Windows binaries.
       - name: Setup Java
-        uses: actions/setup-java@c5195efecf7bdfc987ee8bae7a71cb8b11521c00 # v4.7.1
+        uses: actions/setup-java@dded0888837ed1f317902acf8a20df0ad188d165 # v5.0.0
         with:
           distribution: "zulu"
           java-version: "11.0"
@@ -286,17 +317,17 @@ jobs:
       # Setup GCloud for signing Windows binaries.
       - name: Authenticate to Google Cloud
         id: gcloud_auth
-        uses: google-github-actions/auth@ba79af03959ebeac9769e648f473a284504d9193 # v2.1.10
+        uses: google-github-actions/auth@7c6bc770dae815cd3e89ee6cdf493a5fab2cc093 # v3.0.0
         with:
           workload_identity_provider: ${{ vars.GCP_CODE_SIGNING_WORKLOAD_ID_PROVIDER }}
           service_account: ${{ vars.GCP_CODE_SIGNING_SERVICE_ACCOUNT }}
           token_format: "access_token"
 
       - name: Setup GCloud SDK
-        uses: google-github-actions/setup-gcloud@77e7a554d41e2ee56fc945c52dfd3f33d12def9a # v2.1.4
+        uses: google-github-actions/setup-gcloud@aa5489c8933f4cc7a4f7d45035b3b1440c9c10db # v3.0.1
 
       - name: Download dylibs
-        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
+        uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           name: dylibs
           path: ./build
@@ -350,9 +381,9 @@ jobs:
           set -euo pipefail
           if [[ "${CODER_RELEASE:-}" != *t* ]] || [[ "${CODER_DRY_RUN:-}" == *t* ]]; then
             # Empty value means use the default and avoid building a fresh one.
-            echo "tag=" >> $GITHUB_OUTPUT
+            echo "tag=" >> "$GITHUB_OUTPUT"
           else
-            echo "tag=$(CODER_IMAGE_BASE=ghcr.io/coder/coder-base ./scripts/image_tag.sh)" >> $GITHUB_OUTPUT
+            echo "tag=$(CODER_IMAGE_BASE=ghcr.io/coder/coder-base ./scripts/image_tag.sh)" >> "$GITHUB_OUTPUT"
           fi
 
       - name: Create empty base-build-context directory
@@ -366,7 +397,7 @@ jobs:
       # This uses OIDC authentication, so no auth variables are required.
       - name: Build base Docker image via depot.dev
         if: steps.image-base-tag.outputs.tag != ''
-        uses: depot/build-push-action@2583627a84956d07561420dcc1d0eb1f2af3fac0 # v1.15.0
+        uses: depot/build-push-action@9785b135c3c76c33db102e45be96a25ab55cd507 # v1.16.2
         with:
           project: wl5hnrrkns
           context: base-build-context
@@ -387,7 +418,7 @@ jobs:
           # available immediately
           for i in {1..10}; do
             rc=0
-            raw_manifests=$(docker buildx imagetools inspect --raw "${{ steps.image-base-tag.outputs.tag }}") || rc=$?
+            raw_manifests=$(docker buildx imagetools inspect --raw "${IMAGE_TAG}") || rc=$?
             if [[ "$rc" -eq 0 ]]; then
               break
             fi
@@ -409,6 +440,8 @@ jobs:
           echo "$manifests" | grep -q linux/amd64
           echo "$manifests" | grep -q linux/arm64
           echo "$manifests" | grep -q linux/arm/v7
+        env:
+          IMAGE_TAG: ${{ steps.image-base-tag.outputs.tag }}
 
       # GitHub attestation provides SLSA provenance for Docker images, establishing a verifiable
       # record that these images were built in GitHub Actions with specific inputs and environment.
@@ -421,7 +454,7 @@ jobs:
         id: attest_base
         if: ${{ !inputs.dry_run && steps.image-base-tag.outputs.tag != '' }}
         continue-on-error: true
-        uses: actions/attest@ce27ba3b4a9a139d9a20a4a07d69fabb52f1e5bc # v2.4.0
+        uses: actions/attest@daf44fb950173508f38bd2406030372c1d1162b1 # v3.0.0
         with:
           subject-name: ${{ steps.image-base-tag.outputs.tag }}
           predicate-type: "https://slsa.dev/provenance/v1"
@@ -476,7 +509,7 @@ jobs:
 
           # Save multiarch image tag for attestation
           multiarch_image="$(./scripts/image_tag.sh)"
-          echo "multiarch_image=${multiarch_image}" >> $GITHUB_OUTPUT
+          echo "multiarch_image=${multiarch_image}" >> "$GITHUB_OUTPUT"
 
           # For debugging, print all docker image tags
           docker images
@@ -484,16 +517,15 @@ jobs:
           # if the current version is equal to the highest (according to semver)
           # version in the repo, also create a multi-arch image as ":latest" and
           # push it
-          created_latest_tag=false
           if [[ "$(git tag | grep '^v' | grep -vE '(rc|dev|-|\+|\/)' | sort -r --version-sort | head -n1)" == "v$(./scripts/version.sh)" ]]; then
+            # shellcheck disable=SC2046
             ./scripts/build_docker_multiarch.sh \
               --push \
               --target "$(./scripts/image_tag.sh --version latest)" \
               $(cat build/coder_"$version"_linux_{amd64,arm64,armv7}.tag)
-            created_latest_tag=true
-            echo "created_latest_tag=true" >> $GITHUB_OUTPUT
+            echo "created_latest_tag=true" >> "$GITHUB_OUTPUT"
           else
-            echo "created_latest_tag=false" >> $GITHUB_OUTPUT
+            echo "created_latest_tag=false" >> "$GITHUB_OUTPUT"
           fi
         env:
           CODER_BASE_IMAGE_TAG: ${{ steps.image-base-tag.outputs.tag }}
@@ -501,24 +533,27 @@ jobs:
       - name: SBOM Generation and Attestation
         if: ${{ !inputs.dry_run }}
         env:
-          COSIGN_EXPERIMENTAL: "1"
+          COSIGN_EXPERIMENTAL: '1'
+          MULTIARCH_IMAGE: ${{ steps.build_docker.outputs.multiarch_image }}
+          VERSION: ${{ steps.version.outputs.version }}
+          CREATED_LATEST_TAG: ${{ steps.build_docker.outputs.created_latest_tag }}
         run: |
           set -euxo pipefail
 
           # Generate SBOM for multi-arch image with version in filename
-          echo "Generating SBOM for multi-arch image: ${{ steps.build_docker.outputs.multiarch_image }}"
-          syft "${{ steps.build_docker.outputs.multiarch_image }}" -o spdx-json > coder_${{ steps.version.outputs.version }}_sbom.spdx.json
+          echo "Generating SBOM for multi-arch image: ${MULTIARCH_IMAGE}"
+          syft "${MULTIARCH_IMAGE}" -o spdx-json > "coder_${VERSION}_sbom.spdx.json"
 
           # Attest SBOM to multi-arch image
-          echo "Attesting SBOM to multi-arch image: ${{ steps.build_docker.outputs.multiarch_image }}"
-          cosign clean --force=true "${{ steps.build_docker.outputs.multiarch_image }}"
+          echo "Attesting SBOM to multi-arch image: ${MULTIARCH_IMAGE}"
+          cosign clean --force=true "${MULTIARCH_IMAGE}"
           cosign attest --type spdxjson \
-            --predicate coder_${{ steps.version.outputs.version }}_sbom.spdx.json \
+            --predicate "coder_${VERSION}_sbom.spdx.json" \
             --yes \
-            "${{ steps.build_docker.outputs.multiarch_image }}"
+            "${MULTIARCH_IMAGE}"
 
           # If latest tag was created, also attest it
-          if [[ "${{ steps.build_docker.outputs.created_latest_tag }}" == "true" ]]; then
+          if [[ "${CREATED_LATEST_TAG}" == "true" ]]; then
             latest_tag="$(./scripts/image_tag.sh --version latest)"
             echo "Generating SBOM for latest image: ${latest_tag}"
             syft "${latest_tag}" -o spdx-json > coder_latest_sbom.spdx.json
@@ -535,7 +570,7 @@ jobs:
         id: attest_main
         if: ${{ !inputs.dry_run }}
         continue-on-error: true
-        uses: actions/attest@ce27ba3b4a9a139d9a20a4a07d69fabb52f1e5bc # v2.4.0
+        uses: actions/attest@daf44fb950173508f38bd2406030372c1d1162b1 # v3.0.0
         with:
           subject-name: ${{ steps.build_docker.outputs.multiarch_image }}
           predicate-type: "https://slsa.dev/provenance/v1"
@@ -572,14 +607,14 @@ jobs:
       - name: Get latest tag name
         id: latest_tag
         if: ${{ !inputs.dry_run && steps.build_docker.outputs.created_latest_tag == 'true' }}
-        run: echo "tag=$(./scripts/image_tag.sh --version latest)" >> $GITHUB_OUTPUT
+        run: echo "tag=$(./scripts/image_tag.sh --version latest)" >> "$GITHUB_OUTPUT"
 
       # If this is the highest version according to semver, also attest the "latest" tag
       - name: GitHub Attestation for "latest" Docker image
         id: attest_latest
         if: ${{ !inputs.dry_run && steps.build_docker.outputs.created_latest_tag == 'true' }}
         continue-on-error: true
-        uses: actions/attest@ce27ba3b4a9a139d9a20a4a07d69fabb52f1e5bc # v2.4.0
+        uses: actions/attest@daf44fb950173508f38bd2406030372c1d1162b1 # v3.0.0
         with:
           subject-name: ${{ steps.latest_tag.outputs.tag }}
           predicate-type: "https://slsa.dev/provenance/v1"
@@ -615,7 +650,7 @@ jobs:
       # Report attestation failures but don't fail the workflow
       - name: Check attestation status
         if: ${{ !inputs.dry_run }}
-        run: |
+        run: | # zizmor: ignore[template-injection] We're just reading steps.attest_x.outcome here, no risk of injection
           if [[ "${{ steps.attest_base.outcome }}" == "failure" && "${{ steps.attest_base.conclusion }}" != "skipped" ]]; then
             echo "::warning::GitHub attestation for base image failed"
           fi
@@ -635,7 +670,7 @@ jobs:
         run: ls -lh build
 
       - name: Publish Coder CLI binaries and detached signatures to GCS
-        if: ${{ !inputs.dry_run && github.ref == 'refs/heads/main' && github.repository_owner == 'coder'}}
+        if: ${{ !inputs.dry_run }}
         run: |
           set -euxo pipefail
 
@@ -680,11 +715,11 @@ jobs:
             ./build/*.apk
             ./build/*.deb
             ./build/*.rpm
-            ./coder_${{ steps.version.outputs.version }}_sbom.spdx.json
+            "./coder_${VERSION}_sbom.spdx.json"
           )
 
           # Only include the latest SBOM file if it was created
-          if [[ "${{ steps.build_docker.outputs.created_latest_tag }}" == "true" ]]; then
+          if [[ "${CREATED_LATEST_TAG}" == "true" ]]; then
             files+=(./coder_latest_sbom.spdx.json)
           fi
 
@@ -695,15 +730,17 @@ jobs:
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
           CODER_GPG_RELEASE_KEY_BASE64: ${{ secrets.GPG_RELEASE_KEY_BASE64 }}
+          VERSION: ${{ steps.version.outputs.version }}
+          CREATED_LATEST_TAG: ${{ steps.build_docker.outputs.created_latest_tag }}
 
       - name: Authenticate to Google Cloud
-        uses: google-github-actions/auth@ba79af03959ebeac9769e648f473a284504d9193 # v2.1.10
+        uses: google-github-actions/auth@7c6bc770dae815cd3e89ee6cdf493a5fab2cc093 # v3.0.0
         with:
           workload_identity_provider: ${{ vars.GCP_WORKLOAD_ID_PROVIDER }}
           service_account: ${{ vars.GCP_SERVICE_ACCOUNT }}
 
       - name: Setup GCloud SDK
-        uses: google-github-actions/setup-gcloud@77e7a554d41e2ee56fc945c52dfd3f33d12def9a # 2.1.4
+        uses: google-github-actions/setup-gcloud@aa5489c8933f4cc7a4f7d45035b3b1440c9c10db # 3.0.1
 
       - name: Publish Helm Chart
         if: ${{ !inputs.dry_run }}
@@ -715,14 +752,16 @@ jobs:
           cp "build/provisioner_helm_${version}.tgz" build/helm
           gsutil cp gs://helm.coder.com/v2/index.yaml build/helm/index.yaml
           helm repo index build/helm --url https://helm.coder.com/v2 --merge build/helm/index.yaml
-          gsutil -h "Cache-Control:no-cache,max-age=0" cp build/helm/coder_helm_${version}.tgz gs://helm.coder.com/v2
-          gsutil -h "Cache-Control:no-cache,max-age=0" cp build/helm/provisioner_helm_${version}.tgz gs://helm.coder.com/v2
-          gsutil -h "Cache-Control:no-cache,max-age=0" cp build/helm/index.yaml gs://helm.coder.com/v2
-          gsutil -h "Cache-Control:no-cache,max-age=0" cp helm/artifacthub-repo.yml gs://helm.coder.com/v2
+          gsutil -h "Cache-Control:no-cache,max-age=0" cp "build/helm/coder_helm_${version}.tgz" gs://helm.coder.com/v2
+          gsutil -h "Cache-Control:no-cache,max-age=0" cp "build/helm/provisioner_helm_${version}.tgz" gs://helm.coder.com/v2
+          gsutil -h "Cache-Control:no-cache,max-age=0" cp "build/helm/index.yaml" gs://helm.coder.com/v2
+          gsutil -h "Cache-Control:no-cache,max-age=0" cp "helm/artifacthub-repo.yml" gs://helm.coder.com/v2
+          helm push "build/coder_helm_${version}.tgz" oci://ghcr.io/coder/chart
+          helm push "build/provisioner_helm_${version}.tgz" oci://ghcr.io/coder/chart
 
       - name: Upload artifacts to actions (if dry-run)
         if: ${{ inputs.dry_run }}
-        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: release-artifacts
           path: |
@@ -738,7 +777,7 @@ jobs:
 
       - name: Upload latest sbom artifact to actions (if dry-run)
         if: inputs.dry_run && steps.build_docker.outputs.created_latest_tag == 'true'
-        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: latest-sbom-artifact
           path: ./coder_latest_sbom.spdx.json
@@ -746,7 +785,7 @@ jobs:
 
       - name: Send repository-dispatch event
         if: ${{ !inputs.dry_run }}
-        uses: peter-evans/repository-dispatch@ff45666b9427631e3450c54a1bcbee4d9ff4d7c0 # v3.0.0
+        uses: peter-evans/repository-dispatch@28959ce8df70de7be546dd1250a005dd32156697 # v4.0.1
         with:
           token: ${{ secrets.CDRCI_GITHUB_TOKEN }}
           repository: coder/packages
@@ -763,18 +802,18 @@ jobs:
       # TODO: skip this if it's not a new release (i.e. a backport). This is
       #       fine right now because it just makes a PR that we can close.
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Update homebrew
         env:
-          # Variables used by the `gh` command
           GH_REPO: coder/homebrew-coder
           GH_TOKEN: ${{ secrets.CDRCI_GITHUB_TOKEN }}
+          VERSION: ${{ needs.release.outputs.version }}
         run: |
           # Keep version number around for reference, removing any potential leading v
-          coder_version="$(echo "${{ needs.release.outputs.version }}" | tr -d v)"
+          coder_version="$(echo "${VERSION}" | tr -d v)"
 
           set -euxo pipefail
 
@@ -793,9 +832,9 @@ jobs:
           wget "$checksums_url" -O checksums.txt
 
           # Get the SHAs
-          darwin_arm_sha="$(cat checksums.txt | grep "darwin_arm64.zip" | awk '{ print $1 }')"
-          darwin_intel_sha="$(cat checksums.txt | grep "darwin_amd64.zip" | awk '{ print $1 }')"
-          linux_sha="$(cat checksums.txt | grep "linux_amd64.tar.gz" | awk '{ print $1 }')"
+          darwin_arm_sha="$(grep "darwin_arm64.zip" checksums.txt | awk '{ print $1 }')"
+          darwin_intel_sha="$(grep "darwin_amd64.zip" checksums.txt | awk '{ print $1 }')"
+          linux_sha="$(grep "linux_amd64.tar.gz" checksums.txt | awk '{ print $1 }')"
 
           echo "macOS arm64: $darwin_arm_sha"
           echo "macOS amd64: $darwin_intel_sha"
@@ -808,7 +847,7 @@ jobs:
 
           # Check if a PR already exists.
           pr_count="$(gh pr list --search "head:$brew_branch" --json id,closed | jq -r ".[] | select(.closed == false) | .id" | wc -l)"
-          if [[ "$pr_count" > 0 ]]; then
+          if [ "$pr_count" -gt 0 ]; then
             echo "Bailing out as PR already exists" 2>&1
             exit 0
           fi
@@ -827,8 +866,8 @@ jobs:
             -B master -H "$brew_branch" \
             -t "coder $coder_version" \
             -b "" \
-            -r "${{ github.actor }}" \
-            -a "${{ github.actor }}" \
+            -r "${GITHUB_ACTOR}" \
+            -a "${GITHUB_ACTOR}" \
             -b "This automatic PR was triggered by the release of Coder v$coder_version"
 
   publish-winget:
@@ -839,7 +878,7 @@ jobs:
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
@@ -849,9 +888,10 @@ jobs:
           GH_TOKEN: ${{ secrets.CDRCI_GITHUB_TOKEN }}
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           fetch-depth: 0
+          persist-credentials: false
 
       # If the event that triggered the build was an annotated tag (which our
       # tags are supposed to be), actions/checkout has a bug where the tag in
@@ -870,7 +910,7 @@ jobs:
           # The package version is the same as the tag minus the leading "v".
           # The version in this output already has the leading "v" removed but
           # we do it again to be safe.
-          $version = "${{ needs.release.outputs.version }}".Trim('v')
+          $version = $env:VERSION.Trim('v')
 
           $release_assets = gh release view --repo coder/coder "v${version}" --json assets | `
             ConvertFrom-Json
@@ -902,13 +942,14 @@ jobs:
           # For wingetcreate. We need a real token since we're pushing a commit
           # to GitHub and then making a PR in a different repo.
           WINGET_GH_TOKEN: ${{ secrets.CDRCI_GITHUB_TOKEN }}
+          VERSION: ${{ needs.release.outputs.version }}
 
       - name: Comment on PR
         run: |
           # wait 30 seconds
           Start-Sleep -Seconds 30.0
           # Find the PR that wingetcreate just made.
-          $version = "${{ needs.release.outputs.version }}".Trim('v')
+          $version = $env:VERSION.Trim('v')
           $pr_list = gh pr list --repo microsoft/winget-pkgs --search "author:cdrci Coder.Coder version ${version}" --limit 1 --json number | `
             ConvertFrom-Json
           $pr_number = $pr_list[0].number
@@ -919,6 +960,7 @@ jobs:
           # For gh CLI. We need a real token since we're commenting on a PR in a
           # different repo.
           GH_TOKEN: ${{ secrets.CDRCI_GITHUB_TOKEN }}
+          VERSION: ${{ needs.release.outputs.version }}
 
   # publish-sqlc pushes the latest schema to sqlc cloud.
   # At present these pushes cannot be tagged, so the last push is always the latest.
@@ -929,14 +971,15 @@ jobs:
     if: ${{ !inputs.dry_run }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           fetch-depth: 1
+          persist-credentials: false
 
       # We need golang to run the migration main.go
       - name: Setup Go

.github/workflows/scorecard.yml

@@ -20,17 +20,17 @@ jobs:
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: "Checkout code"
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           persist-credentials: false
 
       - name: "Run analysis"
-        uses: ossf/scorecard-action@05b42c624433fc40578a4040d5cf5e36ddca8cde # v2.4.2
+        uses: ossf/scorecard-action@4eaacf0543bb3f2c246792bd56e8cdeffafb205a # v2.4.3
         with:
           results_file: results.sarif
           results_format: sarif
@@ -39,7 +39,7 @@ jobs:
 
       # Upload the results as artifacts.
       - name: "Upload artifact"
-        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: SARIF file
           path: results.sarif
@@ -47,6 +47,6 @@ jobs:
 
       # Upload the results to GitHub's code scanning dashboard.
       - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@ce28f5bb42b7a9f2c824e633a3f6ee835bab6858 # v3.29.0
+        uses: github/codeql-action/upload-sarif@fe4161a26a8629af62121b670040955b330f9af2 # v3.29.5
         with:
           sarif_file: results.sarif

.github/workflows/security.yaml

@@ -27,18 +27,20 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Setup Go
         uses: ./.github/actions/setup-go
 
       - name: Initialize CodeQL
-        uses: github/codeql-action/init@ce28f5bb42b7a9f2c824e633a3f6ee835bab6858 # v3.29.0
+        uses: github/codeql-action/init@fe4161a26a8629af62121b670040955b330f9af2 # v3.29.5
         with:
           languages: go, javascript
 
@@ -48,7 +50,7 @@ jobs:
           rm Makefile
 
       - name: Perform CodeQL Analysis
-        uses: github/codeql-action/analyze@ce28f5bb42b7a9f2c824e633a3f6ee835bab6858 # v3.29.0
+        uses: github/codeql-action/analyze@fe4161a26a8629af62121b670040955b330f9af2 # v3.29.5
 
       - name: Send Slack notification on failure
         if: ${{ failure() }}
@@ -67,14 +69,15 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
         with:
           fetch-depth: 0
+          persist-credentials: false
 
       - name: Setup Go
         uses: ./.github/actions/setup-go
@@ -134,15 +137,16 @@ jobs:
           # This environment variables forces scripts/build_docker.sh to build
           # the base image tag locally instead of using the cached version from
           # the registry.
-          export CODER_IMAGE_BUILD_BASE_TAG="$(CODER_IMAGE_BASE=coder-base ./scripts/image_tag.sh --version "$version")"
+          CODER_IMAGE_BUILD_BASE_TAG="$(CODER_IMAGE_BASE=coder-base ./scripts/image_tag.sh --version "$version")"
+          export CODER_IMAGE_BUILD_BASE_TAG
 
           # We would like to use make -j here, but it doesn't work with the some recent additions
           # to our code generation.
           make "$image_job"
-          echo "image=$(cat "$image_job")" >> $GITHUB_OUTPUT
+          echo "image=$(cat "$image_job")" >> "$GITHUB_OUTPUT"
 
       - name: Run Trivy vulnerability scanner
-        uses: aquasecurity/trivy-action@76071ef0d7ec797419534a183b498b4d6366cf37
+        uses: aquasecurity/trivy-action@b6643a29fecd7f34b3597bc6acb0a98b03d33ff8
         with:
           image-ref: ${{ steps.build.outputs.image }}
           format: sarif
@@ -150,13 +154,13 @@ jobs:
           severity: "CRITICAL,HIGH"
 
       - name: Upload Trivy scan results to GitHub Security tab
-        uses: github/codeql-action/upload-sarif@ce28f5bb42b7a9f2c824e633a3f6ee835bab6858 # v3.29.0
+        uses: github/codeql-action/upload-sarif@fe4161a26a8629af62121b670040955b330f9af2 # v3.29.5
         with:
           sarif_file: trivy-results.sarif
           category: "Trivy"
 
       - name: Upload Trivy scan results as an artifact
-        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: trivy
           path: trivy-results.sarif

.github/workflows/stale.yaml

@@ -18,12 +18,12 @@ jobs:
       pull-requests: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: stale
-        uses: actions/stale@5bef64f19d7facfb25b37b414482c7164d639639 # v9.1.0
+        uses: actions/stale@5f858e3efba33a5ca4407a664cc011ad407f2008 # v10.1.0
         with:
           stale-issue-label: "stale"
           stale-pr-label: "stale"
@@ -44,7 +44,7 @@ jobs:
           # Start with the oldest issues, always.
           ascending: true
       - name: "Close old issues labeled likely-no"
-        uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea # v7.0.1
+        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
         with:
           github-token: ${{ secrets.GITHUB_TOKEN }}
           script: |
@@ -96,12 +96,14 @@ jobs:
       contents: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout repository
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
       - name: Run delete-old-branches-action
         uses: beatlabs/delete-old-branches-action@4eeeb8740ff8b3cb310296ddd6b43c3387734588 # v0.0.11
         with:
@@ -118,12 +120,12 @@ jobs:
       actions: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Delete PR Cleanup workflow runs
-        uses: Mattraks/delete-workflow-runs@39f0bbed25d76b34de5594dceab824811479e5de # v2.0.6
+        uses: Mattraks/delete-workflow-runs@5bf9a1dac5c4d041c029f0a8370ddf0c5cb5aeb7 # v2.1.0
         with:
           token: ${{ github.token }}
           repository: ${{ github.repository }}
@@ -132,7 +134,7 @@ jobs:
           delete_workflow_pattern: pr-cleanup.yaml
 
       - name: Delete PR Deploy workflow skipped runs
-        uses: Mattraks/delete-workflow-runs@39f0bbed25d76b34de5594dceab824811479e5de # v2.0.6
+        uses: Mattraks/delete-workflow-runs@5bf9a1dac5c4d041c029f0a8370ddf0c5cb5aeb7 # v2.1.0
         with:
           token: ${{ github.token }}
           repository: ${{ github.repository }}

.github/workflows/start-workspace.yaml

@@ -19,7 +19,7 @@ jobs:
     timeout-minutes: 5
     steps:
       - name: Start Coder workspace
-        uses: coder/start-workspace-action@35a4608cefc7e8cc56573cae7c3b85304575cb72
+        uses: coder/start-workspace-action@f97a681b4cc7985c9eef9963750c7cc6ebc93a19
         with:
           github-token: ${{ secrets.GITHUB_TOKEN }}
           github-username: >-

.github/workflows/traiage.yaml

@@ -0,0 +1,190 @@
+name: AI Triage Automation
+
+on:
+  issues:
+    types:
+      - labeled
+  workflow_dispatch:
+    inputs:
+      issue_url:
+        description: "GitHub Issue URL to process"
+        required: true
+        type: string
+      template_name:
+        description: "Coder template to use for workspace"
+        required: true
+        default: "coder"
+        type: string
+      template_preset:
+        description: "Template preset to use"
+        required: false
+        default: ""
+        type: string
+      prefix:
+        description: "Prefix for workspace name"
+        required: false
+        default: "traiage"
+        type: string
+
+jobs:
+  traiage:
+    name: Triage GitHub Issue with Claude Code
+    runs-on: ubuntu-latest
+    if: github.event.label.name == 'traiage' || github.event_name == 'workflow_dispatch'
+    timeout-minutes: 30
+    env:
+      CODER_URL: ${{ secrets.TRAIAGE_CODER_URL }}
+      CODER_SESSION_TOKEN: ${{ secrets.TRAIAGE_CODER_SESSION_TOKEN }}
+    permissions:
+      contents: read
+      issues: write
+      actions: write
+
+    steps:
+      # This is only required for testing locally using nektos/act, so leaving commented out.
+      # An alternative is to use a larger or custom image.
+      # - name: Install Github CLI
+      #   id: install-gh
+      #   run: |
+      #     (type -p wget >/dev/null || (sudo apt update && sudo apt install wget -y)) \
+      #     && sudo mkdir -p -m 755 /etc/apt/keyrings \
+      #     && out=$(mktemp) && wget -nv -O$out https://cli.github.com/packages/githubcli-archive-keyring.gpg \
+      #     && cat $out | sudo tee /etc/apt/keyrings/githubcli-archive-keyring.gpg > /dev/null \
+      #     && sudo chmod go+r /etc/apt/keyrings/githubcli-archive-keyring.gpg \
+      #     && sudo mkdir -p -m 755 /etc/apt/sources.list.d \
+      #     && echo "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/githubcli-archive-keyring.gpg] https://cli.github.com/packages stable main" | sudo tee /etc/apt/sources.list.d/github-cli.list > /dev/null \
+      #     && sudo apt update \
+      #     && sudo apt install gh -y
+
+      - name: Determine Inputs
+        id: determine-inputs
+        if: always()
+        env:
+          GITHUB_ACTOR: ${{ github.actor }}
+          GITHUB_EVENT_ISSUE_HTML_URL: ${{ github.event.issue.html_url }}
+          GITHUB_EVENT_NAME: ${{ github.event_name }}
+          GITHUB_EVENT_USER_ID: ${{ github.event.sender.id }}
+          GITHUB_EVENT_USER_LOGIN: ${{ github.event.sender.login }}
+          INPUTS_ISSUE_URL: ${{ inputs.issue_url }}
+          INPUTS_TEMPLATE_NAME: ${{ inputs.template_name || 'coder' }}
+          INPUTS_TEMPLATE_PRESET: ${{ inputs.template_preset || ''}}
+          INPUTS_PREFIX: ${{ inputs.prefix || 'traiage' }}
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          echo "Using template name: ${INPUTS_TEMPLATE_NAME}"
+          echo "template_name=${INPUTS_TEMPLATE_NAME}" >> "${GITHUB_OUTPUT}"
+
+          echo "Using template preset: ${INPUTS_TEMPLATE_PRESET}"
+          echo "template_preset=${INPUTS_TEMPLATE_PRESET}" >> "${GITHUB_OUTPUT}"
+
+          echo "Using prefix: ${INPUTS_PREFIX}"
+          echo "prefix=${INPUTS_PREFIX}" >> "${GITHUB_OUTPUT}"
+
+          # For workflow_dispatch, use the actor who triggered it
+          # For issues events, use the issue author.
+          if [[ "${GITHUB_EVENT_NAME}" == "workflow_dispatch" ]]; then
+            if ! GITHUB_USER_ID=$(gh api "users/${GITHUB_ACTOR}" --jq '.id'); then
+              echo "::error::Failed to get GitHub user ID for actor ${GITHUB_ACTOR}"
+              exit 1
+            fi
+            echo "Using workflow_dispatch actor: ${GITHUB_ACTOR} (ID: ${GITHUB_USER_ID})"
+            echo "github_user_id=${GITHUB_USER_ID}" >> "${GITHUB_OUTPUT}"
+            echo "github_username=${GITHUB_ACTOR}" >> "${GITHUB_OUTPUT}"
+
+            echo "Using issue URL: ${INPUTS_ISSUE_URL}"
+            echo "issue_url=${INPUTS_ISSUE_URL}" >> "${GITHUB_OUTPUT}"
+
+            exit 0
+          elif [[ "${GITHUB_EVENT_NAME}" == "issues" ]]; then
+            GITHUB_USER_ID=${GITHUB_EVENT_USER_ID}
+            echo "Using issue author: ${GITHUB_EVENT_USER_LOGIN} (ID: ${GITHUB_USER_ID})"
+            echo "github_user_id=${GITHUB_USER_ID}" >> "${GITHUB_OUTPUT}"
+            echo "github_username=${GITHUB_EVENT_USER_LOGIN}" >> "${GITHUB_OUTPUT}"
+
+            echo "Using issue URL: ${GITHUB_EVENT_ISSUE_HTML_URL}"
+            echo "issue_url=${GITHUB_EVENT_ISSUE_HTML_URL}" >> "${GITHUB_OUTPUT}"
+
+            exit 0
+          else
+            echo "::error::Unsupported event type: ${GITHUB_EVENT_NAME}"
+            exit 1
+          fi
+
+      - name: Verify push access
+        env:
+          GITHUB_REPOSITORY: ${{ github.repository }}
+          GH_TOKEN: ${{ github.token }}
+          GITHUB_USERNAME: ${{ steps.determine-inputs.outputs.github_username }}
+          GITHUB_USER_ID: ${{ steps.determine-inputs.outputs.github_user_id }}
+        run: |
+          # Query the actor’s permission on this repo
+          can_push="$(gh api "/repos/${GITHUB_REPOSITORY}/collaborators/${GITHUB_USERNAME}/permission" --jq '.user.permissions.push')"
+          if [[ "${can_push}" != "true" ]]; then
+            echo "::error title=Access Denied::${GITHUB_USERNAME} does not have push access to ${GITHUB_REPOSITORY}"
+            exit 1
+          fi
+
+      - name: Extract context key and description from issue
+        id: extract-context
+        env:
+          ISSUE_URL: ${{ steps.determine-inputs.outputs.issue_url }}
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          issue_number="$(gh issue view "${ISSUE_URL}" --json number --jq '.number')"
+          context_key="gh-${issue_number}"
+
+          TASK_PROMPT=$(cat <<EOF
+          Fix ${ISSUE_URL}
+
+            1. Use the gh CLI to read the issue description and comments.
+            2. Think carefully and try to understand the root cause. If the issue is unclear or not well defined, ask me to clarify and provide more information.
+            3. Write a proposed implementation plan to PLAN.md for me to review before starting implementation. Your plan should use TDD and only make the minimal changes necessary to fix the root cause.
+            4. When I approve your plan, start working on it. If you encounter issues with the plan, ask me for clarification and update the plan as required.
+            5. When you have finished implementation according to the plan, commit and push your changes, and create a PR using the gh CLI for me to review.
+
+          EOF
+          )
+
+          echo "context_key=${context_key}" >> "${GITHUB_OUTPUT}"
+          {
+            echo "TASK_PROMPT<<EOF"
+            echo "${TASK_PROMPT}"
+            echo "EOF"
+          } >> "${GITHUB_OUTPUT}"
+
+      - name: Checkout repository
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          fetch-depth: 1
+          path: ./.github/actions/create-task-action
+          persist-credentials: false
+          ref: main
+          repository: coder/create-task-action
+
+      - name: Create Coder Task
+        id: create_task
+        uses: ./.github/actions/create-task-action
+        with:
+          coder-url: ${{ secrets.TRAIAGE_CODER_URL }}
+          coder-token: ${{ secrets.TRAIAGE_CODER_SESSION_TOKEN }}
+          coder-organization: "default"
+          coder-template-name: coder
+          coder-template-preset: ${{ steps.determine-inputs.outputs.template_preset }}
+          coder-task-name-prefix: gh-coder
+          coder-task-prompt: ${{ steps.extract-context.outputs.task_prompt }}
+          github-user-id: ${{ steps.determine-inputs.outputs.github_user_id }}
+          github-token: ${{ github.token }}
+          github-issue-url: ${{ steps.determine-inputs.outputs.issue_url }}
+          comment-on-issue: ${{ startsWith(steps.determine-inputs.outputs.issue_url, format('{0}/{1}', github.server_url, github.repository)) }}
+
+      - name: Write outputs
+        env:
+          TASK_CREATED: ${{ steps.create_task.outputs.task-created }}
+          TASK_NAME: ${{ steps.create_task.outputs.task-name }}
+          TASK_URL: ${{ steps.create_task.outputs.task-url }}
+        run: |
+          {
+            echo "**Task created:** ${TASK_CREATED}"
+            echo "**Task name:**    ${TASK_NAME}"
+            echo "**Task URL**:     ${TASK_URL}"
+          } >> "${GITHUB_STEP_SUMMARY}"

.github/workflows/typos.toml

@@ -1,5 +1,6 @@
 [default]
 extend-ignore-identifiers-re = ["gho_.*"]
+extend-ignore-re = ["(#|//)\\s*spellchecker:ignore-next-line\\n.*"]
 
 [default.extend-identifiers]
 alog = "alog"
@@ -8,6 +9,7 @@ IST = "IST"
 MacOS = "macOS"
 AKS = "AKS"
 O_WRONLY = "O_WRONLY"
+AIBridge = "AI Bridge"
 
 [default.extend-words]
 AKS = "AKS"
@@ -28,6 +30,7 @@ HELO = "HELO"
 LKE = "LKE"
 byt = "byt"
 typ = "typ"
+Inferrable = "Inferrable"
 
 [files]
 extend-exclude = [
@@ -47,5 +50,5 @@ extend-exclude = [
 	"provisioner/terraform/testdata/**",
 	# notifications' golden files confuse the detector because of quoted-printable encoding
 	"coderd/notifications/testdata/**",
-	"agent/agentcontainers/testdata/devcontainercli/**"
+	"agent/agentcontainers/testdata/devcontainercli/**",
 ]

.github/workflows/weekly-docs.yaml

@@ -21,15 +21,17 @@ jobs:
       pull-requests: write # required to post PR review comments by the action
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@002fdce3c6a235733a90a27c80493a3241e56863 # v2.12.1
+        uses: step-security/harden-runner@95d9a5deda9de15063e7595e9719c11c38c90ae2 # v2.13.2
         with:
           egress-policy: audit
 
       - name: Checkout
-        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
+        with:
+          persist-credentials: false
 
       - name: Check Markdown links
-        uses: umbrelladocs/action-linkspector@e2ccef58c4b9eb89cd71ee23a8629744bba75aa6 # v1.3.5
+        uses: umbrelladocs/action-linkspector@652f85bc57bb1e7d4327260decc10aa68f7694c3 # v1.4.0
         id: markdown-link-check
         # checks all markdown files from /docs including all subfolders
         with:
@@ -41,7 +43,10 @@ jobs:
       - name: Send Slack notification
         if: failure() && github.event_name == 'schedule'
         run: |
-          curl -X POST -H 'Content-type: application/json' -d '{"msg":"Broken links found in the documentation. Please check the logs at ${{ env.LOGS_URL }}"}' ${{ secrets.DOCS_LINK_SLACK_WEBHOOK }}
+          curl \
+            -X POST \
+            -H 'Content-type: application/json' \
+            -d '{"msg":"Broken links found in the documentation. Please check the logs at '"${LOGS_URL}"'"}' "${{ secrets.DOCS_LINK_SLACK_WEBHOOK }}"
           echo "Sent Slack notification"
         env:
           LOGS_URL: https://github.com/coder/coder/actions/runs/${{ github.run_id }}

.github/zizmor.yml

@@ -0,0 +1,4 @@
+rules:
+  cache-poisoning:
+    ignore:
+      - "ci.yaml:184"

.gitignore

@@ -12,6 +12,9 @@ node_modules/
 vendor/
 yarn-error.log
 
+# Test output files
+test-output/
+
 # VSCode settings.
 **/.vscode/*
 # Allow VSCode recommendations and default settings in project root.
@@ -86,3 +89,11 @@ result
 __debug_bin*
 
 **/.claude/settings.local.json
+
+# Local agent configuration
+AGENTS.local.md
+
+/.env
+
+# Ignore plans written by AI agents.
+PLAN.md

.golangci.yaml

@@ -169,6 +169,16 @@ linters-settings:
       - name: var-declaration
       - name: var-naming
       - name: waitgroup-by-value
+  usetesting:
+    # Only os-setenv is enabled because we migrated to usetesting from another linter that
+    # only covered os-setenv.
+    os-setenv: true
+    os-create-temp: false
+    os-mkdir-temp: false
+    os-temp-dir: false
+    os-chdir: false
+    context-background: false
+    context-todo: false
 
   # irrelevant as of Go v1.22: https://go.dev/blog/loopvar-preview
   govet:
@@ -181,7 +191,6 @@ linters-settings:
 
 issues:
   exclude-dirs:
-    - coderd/database/dbmem
     - node_modules
     - .git
 
@@ -253,7 +262,6 @@ linters:
     # - wastedassign
 
     - staticcheck
-    - tenv
     # In Go, it's possible for a package to test it's internal functionality
     # without testing any exported functions. This is enabled to promote
     # decomposing a package before testing it's internals. A function caller
@@ -266,4 +274,5 @@ linters:
     - typecheck
     - unconvert
     - unused
+    - usetesting
     - dupl

.mcp.json

@@ -0,0 +1,36 @@
+{
+  "mcpServers": {
+    "go-language-server": {
+      "type": "stdio",
+      "command": "go",
+      "args": [
+        "run",
+        "github.com/isaacphi/mcp-language-server@latest",
+        "-workspace",
+        "./",
+        "-lsp",
+        "go",
+        "--",
+        "run",
+        "golang.org/x/tools/gopls@latest"
+      ],
+      "env": {}
+    },
+    "typescript-language-server": {
+      "type": "stdio",
+      "command": "go",
+      "args": [
+        "run",
+        "github.com/isaacphi/mcp-language-server@latest",
+        "-workspace",
+        "./site/",
+        "-lsp",
+        "pnpx",
+        "--",
+        "typescript-language-server",
+        "--stdio"
+      ],
+      "env": {}
+    }
+  }
+}

.vscode/settings.json

@@ -49,16 +49,18 @@
 	"[javascript][javascriptreact][json][jsonc][typescript][typescriptreact]": {
 		"editor.defaultFormatter": "biomejs.biome",
 		"editor.codeActionsOnSave": {
-			"quickfix.biome": "explicit"
+			"source.fixAll.biome": "explicit"
 			//   "source.organizeImports.biome": "explicit"
 		}
 	},
 
+	"tailwindCSS.classFunctions": ["cva", "cn"],
 	"[css][html][markdown][yaml]": {
 		"editor.defaultFormatter": "esbenp.prettier-vscode"
 	},
 	"typos.config": ".github/workflows/typos.toml",
 	"[markdown]": {
 		"editor.defaultFormatter": "DavidAnson.vscode-markdownlint"
-	}
+	},
+	"biome.lsp.bin": "site/node_modules/.bin/biome"
 }

AGENTS.md

@@ -0,0 +1 @@
+CLAUDE.md

CLAUDE.md

@@ -1,106 +1,199 @@
 # Coder Development Guidelines
 
-Read [cursor rules](.cursorrules).
+You are an experienced, pragmatic software engineer. You don't over-engineer a solution when a simple one is possible.
+Rule #1: If you want exception to ANY rule, YOU MUST STOP and get explicit permission first. BREAKING THE LETTER OR SPIRIT OF THE RULES IS FAILURE.
 
-## Build/Test/Lint Commands
+## Foundational rules
 
-### Main Commands
+- Doing it right is better than doing it fast. You are not in a rush. NEVER skip steps or take shortcuts.
+- Tedious, systematic work is often the correct solution. Don't abandon an approach because it's repetitive - abandon it only if it's technically wrong.
+- Honesty is a core value.
 
-- `make build` or `make build-fat` - Build all "fat" binaries (includes "server" functionality)
-- `make build-slim` - Build "slim" binaries
-- `make test` - Run Go tests
-- `make test RUN=TestFunctionName` or `go test -v ./path/to/package -run TestFunctionName` - Test single
-- `make test-postgres` - Run tests with Postgres database
-- `make test-race` - Run tests with Go race detector
-- `make test-e2e` - Run end-to-end tests
-- `make lint` - Run all linters
-- `make fmt` - Format all code
-- `make gen` - Generates mocks, database queries and other auto-generated files
+## Our relationship
 
-### Frontend Commands (site directory)
+- Act as a critical peer reviewer. Your job is to disagree with me when I'm wrong, not to please me. Prioritize accuracy and reasoning over agreement.
+- YOU MUST speak up immediately when you don't know something or we're in over our heads
+- YOU MUST call out bad ideas, unreasonable expectations, and mistakes - I depend on this
+- NEVER be agreeable just to be nice - I NEED your HONEST technical judgment
+- NEVER write the phrase "You're absolutely right!"  You are not a sycophant. We're working together because I value your opinion. Do not agree with me unless you can justify it with evidence or reasoning.
+- YOU MUST ALWAYS STOP and ask for clarification rather than making assumptions.
+- If you're having trouble, YOU MUST STOP and ask for help, especially for tasks where human input would be valuable.
+- When you disagree with my approach, YOU MUST push back. Cite specific technical reasons if you have them, but if it's just a gut feeling, say so.
+- If you're uncomfortable pushing back out loud, just say "Houston, we have a problem". I'll know what you mean
+- We discuss architectutral decisions (framework changes, major refactoring, system design) together before implementation. Routine fixes and clear implementations don't need discussion.
 
-- `pnpm build` - Build frontend
-- `pnpm dev` - Run development server
-- `pnpm check` - Run code checks
-- `pnpm format` - Format frontend code
-- `pnpm lint` - Lint frontend code
-- `pnpm test` - Run frontend tests
+## Proactiveness
 
-## Code Style Guidelines
+When asked to do something, just do it - including obvious follow-up actions needed to complete the task properly.
+Only pause to ask for confirmation when:
 
-### Go
+- Multiple valid approaches exist and the choice matters
+- The action would delete or significantly restructure existing code
+- You genuinely don't understand what's being asked
+- Your partner asked a question (answer the question, don't jump to implementation)
 
-- Follow [Effective Go](https://go.dev/doc/effective_go) and [Go's Code Review Comments](https://github.com/golang/go/wiki/CodeReviewComments)
-- Use `gofumpt` for formatting
-- Create packages when used during implementation
-- Validate abstractions against implementations
+@.claude/docs/WORKFLOWS.md
+@package.json
 
-### Error Handling
+## Essential Commands
 
-- Use descriptive error messages
-- Wrap errors with context
-- Propagate errors appropriately
-- Use proper error types
-- (`xerrors.Errorf("failed to X: %w", err)`)
+| Task              | Command                  | Notes                            |
+|-------------------|--------------------------|----------------------------------|
+| **Development**   | `./scripts/develop.sh`   | ⚠️ Don't use manual build        |
+| **Build**         | `make build`             | Fat binaries (includes server)   |
+| **Build Slim**    | `make build-slim`        | Slim binaries                    |
+| **Test**          | `make test`              | Full test suite                  |
+| **Test Single**   | `make test RUN=TestName` | Faster than full suite           |
+| **Test Postgres** | `make test-postgres`     | Run tests with Postgres database |
+| **Test Race**     | `make test-race`         | Run tests with Go race detector  |
+| **Lint**          | `make lint`              | Always run after changes         |
+| **Generate**      | `make gen`               | After database changes           |
+| **Format**        | `make fmt`               | Auto-format code                 |
+| **Clean**         | `make clean`             | Clean build artifacts            |
 
-### Naming
+### Documentation Commands
 
-- Use clear, descriptive names
-- Abbreviate only when obvious
-- Follow Go and TypeScript naming conventions
+- `pnpm run format-docs` - Format markdown tables in docs
+- `pnpm run lint-docs` - Lint and fix markdown files
+- `pnpm run storybook` - Run Storybook (from site directory)
 
-### Comments
+## Critical Patterns
 
-- Document exported functions, types, and non-obvious logic
-- Follow JSDoc format for TypeScript
-- Use godoc format for Go code
+### Database Changes (ALWAYS FOLLOW)
 
-## Commit Style
+1. Modify `coderd/database/queries/*.sql` files
+2. Run `make gen`
+3. If audit errors: update `enterprise/audit/table.go`
+4. Run `make gen` again
 
-- Follow [Conventional Commits 1.0.0](https://www.conventionalcommits.org/en/v1.0.0/)
-- Format: `type(scope): message`
-- Types: `feat`, `fix`, `docs`, `style`, `refactor`, `test`, `chore`
-- Keep message titles concise (~70 characters)
-- Use imperative, present tense in commit titles
+### LSP Navigation (USE FIRST)
 
-## Database queries
+#### Go LSP (for backend code)
 
-- MUST DO! Any changes to database - adding queries, modifying queries should be done in the  `coderd\database\queries\*.sql` files. Use `make gen` to generate necessary changes after.
-- MUST DO! Queries are grouped in files relating to context - e.g. `prebuilds.sql`, `users.sql`, `provisionerjobs.sql`.
-- After making changes to any `coderd\database\queries\*.sql` files you must run `make gen` to generate respective ORM changes.
+- **Find definitions**: `mcp__go-language-server__definition symbolName`
+- **Find references**: `mcp__go-language-server__references symbolName`
+- **Get type info**: `mcp__go-language-server__hover filePath line column`
+- **Rename symbol**: `mcp__go-language-server__rename_symbol filePath line column newName`
+
+#### TypeScript LSP (for frontend code in site/)
+
+- **Find definitions**: `mcp__typescript-language-server__definition symbolName`
+- **Find references**: `mcp__typescript-language-server__references symbolName`
+- **Get type info**: `mcp__typescript-language-server__hover filePath line column`
+- **Rename symbol**: `mcp__typescript-language-server__rename_symbol filePath line column newName`
+
+### OAuth2 Error Handling
+
+```go
+// OAuth2-compliant error responses
+writeOAuth2Error(ctx, rw, http.StatusBadRequest, "invalid_grant", "description")
+```
+
+### Authorization Context
+
+```go
+// Public endpoints needing system access
+app, err := api.Database.GetOAuth2ProviderAppByClientID(dbauthz.AsSystemRestricted(ctx), clientID)
+
+// Authenticated endpoints with user context
+app, err := api.Database.GetOAuth2ProviderAppByClientID(ctx, clientID)
+```
+
+## Quick Reference
+
+### Full workflows available in imported WORKFLOWS.md
+
+### New Feature Checklist
+
+- [ ] Run `git pull` to ensure latest code
+- [ ] Check if feature touches database - you'll need migrations
+- [ ] Check if feature touches audit logs - update `enterprise/audit/table.go`
 
 ## Architecture
 
-### Core Components
+- **coderd**: Main API service
+- **provisionerd**: Infrastructure provisioning
+- **Agents**: Workspace services (SSH, port forwarding)
+- **Database**: PostgreSQL with `dbauthz` authorization
 
-- **coderd**: Main API service connecting workspaces, provisioners, and users
-- **provisionerd**: Execution context for infrastructure-modifying providers
-- **Agents**: Services in remote workspaces providing features like SSH and port forwarding
-- **Workspaces**: Cloud resources defined by Terraform
+## Testing
 
-## Sub-modules
+### Race Condition Prevention
 
-### Template System
+- Use unique identifiers: `fmt.Sprintf("test-client-%s-%d", t.Name(), time.Now().UnixNano())`
+- Never use hardcoded names in concurrent tests
 
-- Templates define infrastructure for workspaces using Terraform
-- Environment variables pass context between Coder and templates
-- Official modules extend development environments
+### OAuth2 Testing
 
-### RBAC System
+- Full suite: `./scripts/oauth2/test-mcp-oauth2.sh`
+- Manual testing: `./scripts/oauth2/test-manual-flow.sh`
 
-- Permissions defined at site, organization, and user levels
-- Object-Action model protects resources
-- Built-in roles: owner, member, auditor, templateAdmin
-- Permission format: `<sign>?<level>.<object>.<id>.<action>`
+### Timing Issues
 
-### Database
+NEVER use `time.Sleep` to mitigate timing issues. If an issue
+seems like it should use `time.Sleep`, read through https://github.com/coder/quartz and specifically the [README](https://github.com/coder/quartz/blob/main/README.md) to better understand how to handle timing issues.
 
-- PostgreSQL 13+ recommended for production
-- Migrations managed with `migrate`
-- Database authorization through `dbauthz` package
+## Code Style
 
-## Frontend
+### Detailed guidelines in imported WORKFLOWS.md
 
-The frontend is contained in the site folder.
+- Follow [Uber Go Style Guide](https://github.com/uber-go/guide/blob/master/style.md)
+- Commit format: `type(scope): message`
 
-For building Frontend refer to [this document](docs/about/contributing/frontend.md)
+### Writing Comments
+
+Code comments should be clear, well-formatted, and add meaningful context.
+
+**Proper sentence structure**: Comments are sentences and should end with
+periods or other appropriate punctuation. This improves readability and
+maintains professional code standards.
+
+**Explain why, not what**: Good comments explain the reasoning behind code
+rather than describing what the code does. The code itself should be
+self-documenting through clear naming and structure. Focus your comments on
+non-obvious decisions, edge cases, or business logic that isn't immediately
+apparent from reading the implementation.
+
+**Line length and wrapping**: Keep comment lines to 80 characters wide
+(including the comment prefix like `//` or `#`). When a comment spans multiple
+lines, wrap it naturally at word boundaries rather than writing one sentence
+per line. This creates more readable, paragraph-like blocks of documentation.
+
+```go
+// Good: Explains the rationale with proper sentence structure.
+// We need a custom timeout here because workspace builds can take several
+// minutes on slow networks, and the default 30s timeout causes false
+// failures during initial template imports.
+ctx, cancel := context.WithTimeout(ctx, 5*time.Minute)
+
+// Bad: Describes what the code does without punctuation or wrapping
+// Set a custom timeout
+// Workspace builds can take a long time
+// Default timeout is too short
+ctx, cancel := context.WithTimeout(ctx, 5*time.Minute)
+```
+
+## Detailed Development Guides
+
+@.claude/docs/ARCHITECTURE.md
+@.claude/docs/OAUTH2.md
+@.claude/docs/TESTING.md
+@.claude/docs/TROUBLESHOOTING.md
+@.claude/docs/DATABASE.md
+
+## Local Configuration
+
+These files may be gitignored, read manually if not auto-loaded.
+
+@AGENTS.local.md
+
+## Common Pitfalls
+
+1. **Audit table errors** → Update `enterprise/audit/table.go`
+2. **OAuth2 errors** → Return RFC-compliant format
+3. **Race conditions** → Use unique test identifiers
+4. **Missing newlines** → Ensure files end with newline
+
+---
+
+*This file stays lean and actionable. Detailed workflows and explanations are imported automatically.*

CODEOWNERS

@@ -1,8 +1,31 @@
-# These APIs are versioned, so any changes need to be carefully reviewed for whether
-# to bump API major or minor versions.
+# These APIs are versioned, so any changes need to be carefully reviewed for
+# whether to bump API major or minor versions.
 agent/proto/ @spikecurtis @johnstcn
+provisionerd/proto/ @spikecurtis @johnstcn
+provisionersdk/proto/ @spikecurtis @johnstcn
 tailnet/proto/ @spikecurtis @johnstcn
 vpn/vpn.proto @spikecurtis @johnstcn
 vpn/version.go @spikecurtis @johnstcn
-provisionerd/proto/ @spikecurtis @johnstcn
-provisionersdk/proto/ @spikecurtis @johnstcn
+
+# This caching code is particularly tricky, and one must be very careful when
+# altering it.
+coderd/files/ @aslilac
+
+coderd/dynamicparameters/ @Emyrk
+coderd/rbac/ @Emyrk
+
+# Mainly dependent on coder/guts, which is maintained by @Emyrk
+scripts/apitypings/ @Emyrk
+scripts/gensite/ @aslilac
+
+# The blood and guts of the autostop algorithm, which is quite complex and
+# requires elite ball knowledge of most of the scheduling code to make changes
+# without inadvertently affecting other parts of the codebase.
+coderd/schedule/autostop.go @deansheather @DanielleMaywood
+
+# Usage tracking code requires intimate knowledge of Tallyman and Metronome, as
+# well as guidance from revenue.
+coderd/usage/ @deansheather @spikecurtis
+enterprise/coderd/usage/ @deansheather @spikecurtis
+
+.github/ 	@jdomeracki-coder

Makefile

@@ -460,16 +460,31 @@ fmt: fmt/ts fmt/go fmt/terraform fmt/shfmt fmt/biome fmt/markdown
 .PHONY: fmt
 
 fmt/go:
+ifdef FILE
+	# Format single file
+	if [[ -f "$(FILE)" ]] && [[ "$(FILE)" == *.go ]] && ! grep -q "DO NOT EDIT" "$(FILE)"; then \
+		echo "$(GREEN)==>$(RESET) $(BOLD)fmt/go$(RESET) $(FILE)"; \
+		go run mvdan.cc/gofumpt@v0.8.0 -w -l "$(FILE)"; \
+	fi
+else
 	go mod tidy
 	echo "$(GREEN)==>$(RESET) $(BOLD)fmt/go$(RESET)"
 	# VS Code users should check out
 	# https://github.com/mvdan/gofumpt#visual-studio-code
 	find . $(FIND_EXCLUSIONS) -type f -name '*.go' -print0 | \
-		xargs -0 grep --null -L "DO NOT EDIT" | \
-		xargs -0 go run mvdan.cc/gofumpt@v0.4.0 -w -l
+		xargs -0 grep -E --null -L '^// Code generated .* DO NOT EDIT\.$$' | \
+		xargs -0 go run mvdan.cc/gofumpt@v0.8.0 -w -l
+endif
 .PHONY: fmt/go
 
 fmt/ts: site/node_modules/.installed
+ifdef FILE
+	# Format single TypeScript/JavaScript file
+	if [[ -f "$(FILE)" ]] && [[ "$(FILE)" == *.ts ]] || [[ "$(FILE)" == *.tsx ]] || [[ "$(FILE)" == *.js ]] || [[ "$(FILE)" == *.jsx ]]; then \
+		echo "$(GREEN)==>$(RESET) $(BOLD)fmt/ts$(RESET) $(FILE)"; \
+		(cd site/ && pnpm exec biome format --write "../$(FILE)"); \
+	fi
+else
 	echo "$(GREEN)==>$(RESET) $(BOLD)fmt/ts$(RESET)"
 	cd site
 # Avoid writing files in CI to reduce file write activity
@@ -478,9 +493,17 @@ ifdef CI
 else
 	pnpm run check:fix
 endif
+endif
 .PHONY: fmt/ts
 
 fmt/biome: site/node_modules/.installed
+ifdef FILE
+	# Format single file with biome
+	if [[ -f "$(FILE)" ]] && [[ "$(FILE)" == *.ts ]] || [[ "$(FILE)" == *.tsx ]] || [[ "$(FILE)" == *.js ]] || [[ "$(FILE)" == *.jsx ]]; then \
+		echo "$(GREEN)==>$(RESET) $(BOLD)fmt/biome$(RESET) $(FILE)"; \
+		(cd site/ && pnpm exec biome format --write "../$(FILE)"); \
+	fi
+else
 	echo "$(GREEN)==>$(RESET) $(BOLD)fmt/biome$(RESET)"
 	cd site/
 # Avoid writing files in CI to reduce file write activity
@@ -489,14 +512,30 @@ ifdef CI
 else
 	pnpm run format
 endif
+endif
 .PHONY: fmt/biome
 
 fmt/terraform: $(wildcard *.tf)
+ifdef FILE
+	# Format single Terraform file
+	if [[ -f "$(FILE)" ]] && [[ "$(FILE)" == *.tf ]] || [[ "$(FILE)" == *.tfvars ]]; then \
+		echo "$(GREEN)==>$(RESET) $(BOLD)fmt/terraform$(RESET) $(FILE)"; \
+		terraform fmt "$(FILE)"; \
+	fi
+else
 	echo "$(GREEN)==>$(RESET) $(BOLD)fmt/terraform$(RESET)"
 	terraform fmt -recursive
+endif
 .PHONY: fmt/terraform
 
 fmt/shfmt: $(SHELL_SRC_FILES)
+ifdef FILE
+	# Format single shell script
+	if [[ -f "$(FILE)" ]] && [[ "$(FILE)" == *.sh ]]; then \
+		echo "$(GREEN)==>$(RESET) $(BOLD)fmt/shfmt$(RESET) $(FILE)"; \
+		shfmt -w "$(FILE)"; \
+	fi
+else
 	echo "$(GREEN)==>$(RESET) $(BOLD)fmt/shfmt$(RESET)"
 # Only do diff check in CI, errors on diff.
 ifdef CI
@@ -504,14 +543,25 @@ ifdef CI
 else
 	shfmt -w $(SHELL_SRC_FILES)
 endif
+endif
 .PHONY: fmt/shfmt
 
 fmt/markdown: node_modules/.installed
+ifdef FILE
+	# Format single markdown file
+	if [[ -f "$(FILE)" ]] && [[ "$(FILE)" == *.md ]]; then \
+		echo "$(GREEN)==>$(RESET) $(BOLD)fmt/markdown$(RESET) $(FILE)"; \
+		pnpm exec markdown-table-formatter "$(FILE)"; \
+	fi
+else
 	echo "$(GREEN)==>$(RESET) $(BOLD)fmt/markdown$(RESET)"
 	pnpm format-docs
+endif
 .PHONY: fmt/markdown
 
-lint: lint/shellcheck lint/go lint/ts lint/examples lint/helm lint/site-icons lint/markdown
+# Note: we don't run zizmor in the lint target because it takes a while. CI
+#       runs it explicitly.
+lint: lint/shellcheck lint/go lint/ts lint/examples lint/helm lint/site-icons lint/markdown lint/actions/actionlint lint/check-scopes
 .PHONY: lint
 
 lint/site-icons:
@@ -528,6 +578,7 @@ lint/go:
 	./scripts/check_codersdk_imports.sh
 	linter_ver=$(shell egrep -o 'GOLANGCI_LINT_VERSION=\S+' dogfood/coder/Dockerfile | cut -d '=' -f 2)
 	go run github.com/golangci/golangci-lint/cmd/golangci-lint@v$$linter_ver run
+	go run github.com/coder/paralleltestctx/cmd/paralleltestctx@v0.0.1 -custom-funcs="testutil.Context" ./...
 .PHONY: lint/go
 
 lint/examples:
@@ -549,13 +600,31 @@ lint/markdown: node_modules/.installed
 	pnpm lint-docs
 .PHONY: lint/markdown
 
+lint/actions: lint/actions/actionlint lint/actions/zizmor
+.PHONY: lint/actions
+
+lint/actions/actionlint:
+	go run github.com/rhysd/actionlint/cmd/actionlint@v1.7.7
+.PHONY: lint/actions/actionlint
+
+lint/actions/zizmor:
+	./scripts/zizmor.sh \
+		--strict-collection \
+		--persona=regular \
+		.
+.PHONY: lint/actions/zizmor
+
+# Verify api_key_scope enum contains all RBAC <resource>:<action> values.
+lint/check-scopes: coderd/database/dump.sql
+	go run ./scripts/check-scopes
+.PHONY: lint/check-scopes
+
 # All files generated by the database should be added here, and this can be used
 # as a target for jobs that need to run after the database is generated.
 DB_GEN_FILES := \
 	coderd/database/dump.sql \
 	coderd/database/querier.go \
 	coderd/database/unique_constraint.go \
-	coderd/database/dbmem/dbmem.go \
 	coderd/database/dbmetrics/dbmetrics.go \
 	coderd/database/dbauthz/dbauthz.go \
 	coderd/database/dbmock/dbmock.go
@@ -566,16 +635,24 @@ TAILNETTEST_MOCKS := \
 	tailnet/tailnettest/workspaceupdatesprovidermock.go \
 	tailnet/tailnettest/subscriptionmock.go
 
+AIBRIDGED_MOCKS := \
+	enterprise/aibridged/aibridgedmock/clientmock.go \
+	enterprise/aibridged/aibridgedmock/poolmock.go
+
 GEN_FILES := \
 	tailnet/proto/tailnet.pb.go \
 	agent/proto/agent.pb.go \
+	agent/agentsocket/proto/agentsocket.pb.go \
 	provisionersdk/proto/provisioner.pb.go \
 	provisionerd/proto/provisionerd.pb.go \
 	vpn/vpn.pb.go \
+	enterprise/aibridged/proto/aibridged.pb.go \
 	$(DB_GEN_FILES) \
 	$(SITE_GEN_FILES) \
 	coderd/rbac/object_gen.go \
 	codersdk/rbacresources_gen.go \
+	coderd/rbac/scopes_constants_gen.go \
+	codersdk/apikey_scopes_gen.go \
 	docs/admin/integrations/prometheus.md \
 	docs/reference/cli/index.md \
 	docs/admin/security/audit-logs.md \
@@ -588,7 +665,9 @@ GEN_FILES := \
 	coderd/database/pubsub/psmock/psmock.go \
 	agent/agentcontainers/acmock/acmock.go \
 	agent/agentcontainers/dcspec/dcspec_gen.go \
-	coderd/httpmw/loggermw/loggermock/loggermock.go
+	coderd/httpmw/loggermw/loggermock/loggermock.go \
+	codersdk/workspacesdk/agentconnmock/agentconnmock.go \
+	$(AIBRIDGED_MOCKS)
 
 # all gen targets should be added here and to gen/mark-fresh
 gen: gen/db gen/golden-files $(GEN_FILES)
@@ -598,6 +677,7 @@ gen/db: $(DB_GEN_FILES)
 .PHONY: gen/db
 
 gen/golden-files: \
+	agent/unit/testdata/.gen-golden \
 	cli/testdata/.gen-golden \
 	coderd/.gen-golden \
 	coderd/notifications/.gen-golden \
@@ -617,12 +697,15 @@ gen/mark-fresh:
 		agent/proto/agent.pb.go \
 		provisionersdk/proto/provisioner.pb.go \
 		provisionerd/proto/provisionerd.pb.go \
+		agent/agentsocket/proto/agentsocket.pb.go \
 		vpn/vpn.pb.go \
+		enterprise/aibridged/proto/aibridged.pb.go \
 		coderd/database/dump.sql \
 		$(DB_GEN_FILES) \
 		site/src/api/typesGenerated.ts \
 		coderd/rbac/object_gen.go \
 		codersdk/rbacresources_gen.go \
+		coderd/rbac/scopes_constants_gen.go \
 		site/src/api/rbacresourcesGenerated.ts \
 		site/src/api/countriesGenerated.ts \
 		docs/admin/integrations/prometheus.md \
@@ -638,6 +721,8 @@ gen/mark-fresh:
 		agent/agentcontainers/acmock/acmock.go \
 		agent/agentcontainers/dcspec/dcspec_gen.go \
 		coderd/httpmw/loggermw/loggermock/loggermock.go \
+		codersdk/workspacesdk/agentconnmock/agentconnmock.go \
+		$(AIBRIDGED_MOCKS) \
 		"
 
 	for file in $$files; do
@@ -681,6 +766,14 @@ coderd/httpmw/loggermw/loggermock/loggermock.go: coderd/httpmw/loggermw/logger.g
 	go generate ./coderd/httpmw/loggermw/loggermock/
 	touch "$@"
 
+codersdk/workspacesdk/agentconnmock/agentconnmock.go: codersdk/workspacesdk/agentconn.go
+	go generate ./codersdk/workspacesdk/agentconnmock/
+	touch "$@"
+
+$(AIBRIDGED_MOCKS): enterprise/aibridged/client.go enterprise/aibridged/pool.go
+	go generate ./enterprise/aibridged/aibridgedmock/
+	touch "$@"
+
 agent/agentcontainers/dcspec/dcspec_gen.go: \
 	node_modules/.installed \
 	agent/agentcontainers/dcspec/devContainer.base.schema.json \
@@ -709,6 +802,14 @@ agent/proto/agent.pb.go: agent/proto/agent.proto
 		--go-drpc_opt=paths=source_relative \
 		./agent/proto/agent.proto
 
+agent/agentsocket/proto/agentsocket.pb.go: agent/agentsocket/proto/agentsocket.proto
+	protoc \
+		--go_out=. \
+		--go_opt=paths=source_relative \
+		--go-drpc_out=. \
+		--go-drpc_opt=paths=source_relative \
+		./agent/agentsocket/proto/agentsocket.proto
+
 provisionersdk/proto/provisioner.pb.go: provisionersdk/proto/provisioner.proto
 	protoc \
 		--go_out=. \
@@ -731,6 +832,14 @@ vpn/vpn.pb.go: vpn/vpn.proto
 		--go_opt=paths=source_relative \
 		./vpn/vpn.proto
 
+enterprise/aibridged/proto/aibridged.pb.go: enterprise/aibridged/proto/aibridged.proto
+	protoc \
+		--go_out=. \
+		--go_opt=paths=source_relative \
+		--go-drpc_out=. \
+		--go-drpc_opt=paths=source_relative \
+		./enterprise/aibridged/proto/aibridged.proto
+
 site/src/api/typesGenerated.ts: site/node_modules/.installed $(wildcard scripts/apitypings/*) $(shell find ./codersdk $(FIND_EXCLUSIONS) -type f -name '*.go')
 	# -C sets the directory for the go run command
 	go run -C ./scripts/apitypings main.go > $@
@@ -757,6 +866,15 @@ coderd/rbac/object_gen.go: scripts/typegen/rbacobject.gotmpl scripts/typegen/mai
 	rmdir -v "$$tempdir"
 	touch "$@"
 
+coderd/rbac/scopes_constants_gen.go: scripts/typegen/scopenames.gotmpl scripts/typegen/main.go coderd/rbac/policy/policy.go
+	# Generate typed low-level ScopeName constants from RBACPermissions
+	# Write to a temp file first to avoid truncating the package during build
+	# since the generator imports the rbac package.
+	tempfile=$(shell mktemp /tmp/scopes_constants_gen.XXXXXX)
+	go run ./scripts/typegen/main.go rbac scopenames > "$$tempfile"
+	mv -v "$$tempfile" coderd/rbac/scopes_constants_gen.go
+	touch "$@"
+
 codersdk/rbacresources_gen.go: scripts/typegen/codersdk.gotmpl scripts/typegen/main.go coderd/rbac/object.go coderd/rbac/policy/policy.go
 	# Do no overwrite codersdk/rbacresources_gen.go directly, as it would make the file empty, breaking
  	# the `codersdk` package and any parallel build targets.
@@ -764,6 +882,12 @@ codersdk/rbacresources_gen.go: scripts/typegen/codersdk.gotmpl scripts/typegen/m
 	mv /tmp/rbacresources_gen.go codersdk/rbacresources_gen.go
 	touch "$@"
 
+codersdk/apikey_scopes_gen.go: scripts/apikeyscopesgen/main.go coderd/rbac/scopes_catalog.go coderd/rbac/scopes.go
+	# Generate SDK constants for external API key scopes.
+	go run ./scripts/apikeyscopesgen > /tmp/apikey_scopes_gen.go
+	mv /tmp/apikey_scopes_gen.go codersdk/apikey_scopes_gen.go
+	touch "$@"
+
 site/src/api/rbacresourcesGenerated.ts: site/node_modules/.installed scripts/typegen/codersdk.gotmpl scripts/typegen/main.go coderd/rbac/object.go coderd/rbac/policy/policy.go
 	go run scripts/typegen/main.go rbac typescript > "$@"
 	(cd site/ && pnpm exec biome format --write src/api/rbacresourcesGenerated.ts)
@@ -839,6 +963,10 @@ clean/golden-files:
 		-type f -name '*.golden' -delete
 .PHONY: clean/golden-files
 
+agent/unit/testdata/.gen-golden: $(wildcard agent/unit/testdata/*.golden) $(GO_SRC_FILES) $(wildcard agent/unit/*_test.go)
+	TZ=UTC go test ./agent/unit -run="TestGraph" -update
+	touch "$@"
+
 cli/testdata/.gen-golden: $(wildcard cli/testdata/*.golden) $(wildcard cli/*.tpl) $(GO_SRC_FILES) $(wildcard cli/*_test.go)
 	TZ=UTC go test ./cli -run="Test(CommandHelp|ServerYAML|ErrorExamples|.*Golden)" -update
 	touch "$@"
@@ -888,12 +1016,31 @@ else
 GOTESTSUM_RETRY_FLAGS :=
 endif
 
+# default to 8x8 parallelism to avoid overwhelming our workspaces. Hopefully we can remove these defaults
+# when we get our test suite's resource utilization under control.
+GOTEST_FLAGS := -v -p $(or $(TEST_NUM_PARALLEL_PACKAGES),"8") -parallel=$(or $(TEST_NUM_PARALLEL_TESTS),"8")
+
+# The most common use is to set TEST_COUNT=1 to avoid Go's test cache.
+ifdef TEST_COUNT
+GOTEST_FLAGS += -count=$(TEST_COUNT)
+endif
+
+ifdef TEST_SHORT
+GOTEST_FLAGS += -short
+endif
+
+ifdef RUN
+GOTEST_FLAGS += -run $(RUN)
+endif
+
+TEST_PACKAGES ?= ./...
+
 test:
-	$(GIT_FLAGS) gotestsum --format standard-quiet $(GOTESTSUM_RETRY_FLAGS) --packages="./..." -- -v -short -count=1 $(if $(RUN),-run $(RUN))
+	$(GIT_FLAGS) gotestsum --format standard-quiet $(GOTESTSUM_RETRY_FLAGS) --packages="$(TEST_PACKAGES)" -- $(GOTEST_FLAGS)
 .PHONY: test
 
 test-cli:
-	$(GIT_FLAGS) gotestsum --format standard-quiet $(GOTESTSUM_RETRY_FLAGS) --packages="./cli/..." -- -v -short -count=1
+	$(MAKE) test TEST_PACKAGES="./cli..."
 .PHONY: test-cli
 
 # sqlc-cloud-is-setup will fail if no SQLc auth token is set. Use this as a
@@ -929,7 +1076,7 @@ sqlc-vet: test-postgres-docker
 test-postgres: test-postgres-docker
 	# The postgres test is prone to failure, so we limit parallelism for
 	# more consistent execution.
-	$(GIT_FLAGS)  DB=ci gotestsum \
+	$(GIT_FLAGS)  gotestsum \
 		--junitfile="gotests.xml" \
 		--jsonfile="gotests.json" \
 		$(GOTESTSUM_RETRY_FLAGS) \
@@ -1045,3 +1192,8 @@ endif
 
 dogfood/coder/nix.hash: flake.nix flake.lock
 	sha256sum flake.nix flake.lock >./dogfood/coder/nix.hash
+
+# Count the number of test databases created per test package.
+count-test-databases:
+	PGPASSWORD=postgres psql -h localhost -U postgres -d coder_testing -P pager=off -c 'SELECT test_package, count(*) as count from test_databases GROUP BY test_package ORDER BY count DESC'
+.PHONY: count-test-databases

agent/agent.go

@@ -8,6 +8,7 @@ import (
 	"fmt"
 	"hash/fnv"
 	"io"
+	"maps"
 	"net"
 	"net/http"
 	"net/netip"
@@ -40,6 +41,7 @@ import (
 	"github.com/coder/coder/v2/agent/agentcontainers"
 	"github.com/coder/coder/v2/agent/agentexec"
 	"github.com/coder/coder/v2/agent/agentscripts"
+	"github.com/coder/coder/v2/agent/agentsocket"
 	"github.com/coder/coder/v2/agent/agentssh"
 	"github.com/coder/coder/v2/agent/proto"
 	"github.com/coder/coder/v2/agent/proto/resourcesmonitor"
@@ -70,17 +72,21 @@ const (
 )
 
 type Options struct {
-	Filesystem                   afero.Fs
-	LogDir                       string
-	TempDir                      string
-	ScriptDataDir                string
-	ExchangeToken                func(ctx context.Context) (string, error)
-	Client                       Client
-	ReconnectingPTYTimeout       time.Duration
-	EnvironmentVariables         map[string]string
-	Logger                       slog.Logger
-	IgnorePorts                  map[int]string
-	PortCacheDuration            time.Duration
+	Filesystem             afero.Fs
+	LogDir                 string
+	TempDir                string
+	ScriptDataDir          string
+	Client                 Client
+	ReconnectingPTYTimeout time.Duration
+	EnvironmentVariables   map[string]string
+	Logger                 slog.Logger
+	// IgnorePorts tells the api handler which ports to ignore when
+	// listing all listening ports. This is helpful to hide ports that
+	// are used by the agent, that the user does not care about.
+	IgnorePorts map[int]string
+	// ListeningPortsGetter is used to get the list of listening ports. Only
+	// tests should set this. If unset, a default that queries the OS will be used.
+	ListeningPortsGetter         ListeningPortsGetter
 	SSHMaxTimeout                time.Duration
 	TailnetListenPort            uint16
 	Subsystems                   []codersdk.AgentSubsystem
@@ -92,13 +98,16 @@ type Options struct {
 	Devcontainers                bool
 	DevcontainerAPIOptions       []agentcontainers.Option // Enable Devcontainers for these to be effective.
 	Clock                        quartz.Clock
+	SocketServerEnabled          bool
+	SocketPath                   string // Path for the agent socket server socket
 }
 
 type Client interface {
 	ConnectRPC26(ctx context.Context) (
 		proto.DRPCAgentClient26, tailnetproto.DRPCTailnetClient26, error,
 	)
-	RewriteDERPMap(derpMap *tailcfg.DERPMap)
+	tailnet.DERPMapRewriter
+	agentsdk.RefreshableSessionTokenProvider
 }
 
 type Agent interface {
@@ -131,20 +140,13 @@ func New(options Options) Agent {
 		}
 		options.ScriptDataDir = options.TempDir
 	}
-	if options.ExchangeToken == nil {
-		options.ExchangeToken = func(_ context.Context) (string, error) {
-			return "", nil
-		}
-	}
 	if options.ReportMetadataInterval == 0 {
 		options.ReportMetadataInterval = time.Second
 	}
 	if options.ServiceBannerRefreshInterval == 0 {
 		options.ServiceBannerRefreshInterval = 2 * time.Minute
 	}
-	if options.PortCacheDuration == 0 {
-		options.PortCacheDuration = 1 * time.Second
-	}
+
 	if options.Clock == nil {
 		options.Clock = quartz.NewReal()
 	}
@@ -158,31 +160,38 @@ func New(options Options) Agent {
 		options.Execer = agentexec.DefaultExecer
 	}
 
+	if options.ListeningPortsGetter == nil {
+		options.ListeningPortsGetter = &osListeningPortsGetter{
+			cacheDuration: 1 * time.Second,
+		}
+	}
+
 	hardCtx, hardCancel := context.WithCancel(context.Background())
 	gracefulCtx, gracefulCancel := context.WithCancel(hardCtx)
 	a := &agent{
-		clock:                              options.Clock,
-		tailnetListenPort:                  options.TailnetListenPort,
-		reconnectingPTYTimeout:             options.ReconnectingPTYTimeout,
-		logger:                             options.Logger,
-		gracefulCtx:                        gracefulCtx,
-		gracefulCancel:                     gracefulCancel,
-		hardCtx:                            hardCtx,
-		hardCancel:                         hardCancel,
-		coordDisconnected:                  make(chan struct{}),
-		environmentVariables:               options.EnvironmentVariables,
-		client:                             options.Client,
-		exchangeToken:                      options.ExchangeToken,
-		filesystem:                         options.Filesystem,
-		logDir:                             options.LogDir,
-		tempDir:                            options.TempDir,
-		scriptDataDir:                      options.ScriptDataDir,
-		lifecycleUpdate:                    make(chan struct{}, 1),
-		lifecycleReported:                  make(chan codersdk.WorkspaceAgentLifecycle, 1),
-		lifecycleStates:                    []agentsdk.PostLifecycleRequest{{State: codersdk.WorkspaceAgentLifecycleCreated}},
-		reportConnectionsUpdate:            make(chan struct{}, 1),
-		ignorePorts:                        options.IgnorePorts,
-		portCacheDuration:                  options.PortCacheDuration,
+		clock:                   options.Clock,
+		tailnetListenPort:       options.TailnetListenPort,
+		reconnectingPTYTimeout:  options.ReconnectingPTYTimeout,
+		logger:                  options.Logger,
+		gracefulCtx:             gracefulCtx,
+		gracefulCancel:          gracefulCancel,
+		hardCtx:                 hardCtx,
+		hardCancel:              hardCancel,
+		coordDisconnected:       make(chan struct{}),
+		environmentVariables:    options.EnvironmentVariables,
+		client:                  options.Client,
+		filesystem:              options.Filesystem,
+		logDir:                  options.LogDir,
+		tempDir:                 options.TempDir,
+		scriptDataDir:           options.ScriptDataDir,
+		lifecycleUpdate:         make(chan struct{}, 1),
+		lifecycleReported:       make(chan codersdk.WorkspaceAgentLifecycle, 1),
+		lifecycleStates:         []agentsdk.PostLifecycleRequest{{State: codersdk.WorkspaceAgentLifecycleCreated}},
+		reportConnectionsUpdate: make(chan struct{}, 1),
+		listeningPortsHandler: listeningPortsHandler{
+			getter:      options.ListeningPortsGetter,
+			ignorePorts: maps.Clone(options.IgnorePorts),
+		},
 		reportMetadataInterval:             options.ReportMetadataInterval,
 		announcementBannersRefreshInterval: options.ServiceBannerRefreshInterval,
 		sshMaxTimeout:                      options.SSHMaxTimeout,
@@ -196,6 +205,8 @@ func New(options Options) Agent {
 
 		devcontainers:       options.Devcontainers,
 		containerAPIOptions: options.DevcontainerAPIOptions,
+		socketPath:          options.SocketPath,
+		socketServerEnabled: options.SocketServerEnabled,
 	}
 	// Initially, we have a closed channel, reflecting the fact that we are not initially connected.
 	// Each time we connect we replace the channel (while holding the closeMutex) with a new one
@@ -203,27 +214,21 @@ func New(options Options) Agent {
 	// coordinator during shut down.
 	close(a.coordDisconnected)
 	a.announcementBanners.Store(new([]codersdk.BannerConfig))
-	a.sessionToken.Store(new(string))
 	a.init()
 	return a
 }
 
 type agent struct {
-	clock             quartz.Clock
-	logger            slog.Logger
-	client            Client
-	exchangeToken     func(ctx context.Context) (string, error)
-	tailnetListenPort uint16
-	filesystem        afero.Fs
-	logDir            string
-	tempDir           string
-	scriptDataDir     string
-	// ignorePorts tells the api handler which ports to ignore when
-	// listing all listening ports. This is helpful to hide ports that
-	// are used by the agent, that the user does not care about.
-	ignorePorts       map[int]string
-	portCacheDuration time.Duration
-	subsystems        []codersdk.AgentSubsystem
+	clock                 quartz.Clock
+	logger                slog.Logger
+	client                Client
+	tailnetListenPort     uint16
+	filesystem            afero.Fs
+	logDir                string
+	tempDir               string
+	scriptDataDir         string
+	listeningPortsHandler listeningPortsHandler
+	subsystems            []codersdk.AgentSubsystem
 
 	reconnectingPTYTimeout time.Duration
 	reconnectingPTYServer  *reconnectingpty.Server
@@ -254,7 +259,6 @@ type agent struct {
 	scriptRunner                       *agentscripts.Runner
 	announcementBanners                atomic.Pointer[[]codersdk.BannerConfig] // announcementBanners is atomic because it is periodically updated.
 	announcementBannersRefreshInterval time.Duration
-	sessionToken                       atomic.Pointer[string]
 	sshServer                          *agentssh.Server
 	sshMaxTimeout                      time.Duration
 	blockFileTransfer                  bool
@@ -280,6 +284,10 @@ type agent struct {
 	devcontainers       bool
 	containerAPIOptions []agentcontainers.Option
 	containerAPI        *agentcontainers.API
+
+	socketServerEnabled bool
+	socketPath          string
+	socketServer        *agentsocket.Server
 }
 
 func (a *agent) TailnetConn() *tailnet.Conn {
@@ -336,18 +344,16 @@ func (a *agent) init() {
 	// will not report anywhere.
 	a.scriptRunner.RegisterMetrics(a.prometheusRegistry)
 
-	if a.devcontainers {
-		containerAPIOpts := []agentcontainers.Option{
-			agentcontainers.WithExecer(a.execer),
-			agentcontainers.WithCommandEnv(a.sshServer.CommandEnv),
-			agentcontainers.WithScriptLogger(func(logSourceID uuid.UUID) agentcontainers.ScriptLogger {
-				return a.logSender.GetScriptLogger(logSourceID)
-			}),
-		}
-		containerAPIOpts = append(containerAPIOpts, a.containerAPIOptions...)
-
-		a.containerAPI = agentcontainers.NewAPI(a.logger.Named("containers"), containerAPIOpts...)
+	containerAPIOpts := []agentcontainers.Option{
+		agentcontainers.WithExecer(a.execer),
+		agentcontainers.WithCommandEnv(a.sshServer.CommandEnv),
+		agentcontainers.WithScriptLogger(func(logSourceID uuid.UUID) agentcontainers.ScriptLogger {
+			return a.logSender.GetScriptLogger(logSourceID)
+		}),
 	}
+	containerAPIOpts = append(containerAPIOpts, a.containerAPIOptions...)
+
+	a.containerAPI = agentcontainers.NewAPI(a.logger.Named("containers"), containerAPIOpts...)
 
 	a.reconnectingPTYServer = reconnectingpty.NewServer(
 		a.logger.Named("reconnecting-pty"),
@@ -361,9 +367,32 @@ func (a *agent) init() {
 			s.ExperimentalContainers = a.devcontainers
 		},
 	)
+
+	a.initSocketServer()
+
 	go a.runLoop()
 }
 
+// initSocketServer initializes server that allows direct communication with a workspace agent using IPC.
+func (a *agent) initSocketServer() {
+	if !a.socketServerEnabled {
+		a.logger.Info(a.hardCtx, "socket server is disabled")
+		return
+	}
+
+	server, err := agentsocket.NewServer(
+		a.logger.Named("socket"),
+		agentsocket.WithPath(a.socketPath),
+	)
+	if err != nil {
+		a.logger.Warn(a.hardCtx, "failed to create socket server", slog.Error(err), slog.F("path", a.socketPath))
+		return
+	}
+
+	a.socketServer = server
+	a.logger.Debug(a.hardCtx, "socket server started", slog.F("path", a.socketPath))
+}
+
 // runLoop attempts to start the agent in a retry loop.
 // Coder may be offline temporarily, a connection issue
 // may be happening, but regardless after the intermittent
@@ -565,7 +594,6 @@ func (a *agent) reportMetadata(ctx context.Context, aAPI proto.DRPCAgentClient26
 			// channel to synchronize the results and avoid both messy
 			// mutex logic and overloading the API.
 			for _, md := range manifest.Metadata {
-				md := md
 				// We send the result to the channel in the goroutine to avoid
 				// sending the same result multiple times. So, we don't care about
 				// the return values.
@@ -793,11 +821,15 @@ func (a *agent) reportConnectionsLoop(ctx context.Context, aAPI proto.DRPCAgentC
 			logger.Debug(ctx, "reporting connection")
 			_, err := aAPI.ReportConnection(ctx, payload)
 			if err != nil {
-				return xerrors.Errorf("failed to report connection: %w", err)
+				// Do not fail the loop if we fail to report a connection, just
+				// log a warning.
+				// Related to https://github.com/coder/coder/issues/20194
+				logger.Warn(ctx, "failed to report connection to server", slog.Error(err))
+				// keep going, we still need to remove it from the slice
+			} else {
+				logger.Debug(ctx, "successfully reported connection")
 			}
 
-			logger.Debug(ctx, "successfully reported connection")
-
 			// Remove the payload we sent.
 			a.reportConnectionsMu.Lock()
 			a.reportConnections[0] = nil // Release the pointer from the underlying array.
@@ -828,6 +860,13 @@ func (a *agent) reportConnection(id uuid.UUID, connectionType proto.Connection_T
 		ip = host
 	}
 
+	// If the IP is "localhost" (which it can be in some cases), set it to
+	// 127.0.0.1 instead.
+	// Related to https://github.com/coder/coder/issues/20194
+	if ip == "localhost" {
+		ip = "127.0.0.1"
+	}
+
 	a.reportConnectionsMu.Lock()
 	defer a.reportConnectionsMu.Unlock()
 
@@ -919,11 +958,10 @@ func (a *agent) run() (retErr error) {
 	// This allows the agent to refresh its token if necessary.
 	// For instance identity this is required, since the instance
 	// may not have re-provisioned, but a new agent ID was created.
-	sessionToken, err := a.exchangeToken(a.hardCtx)
+	err := a.client.RefreshToken(a.hardCtx)
 	if err != nil {
-		return xerrors.Errorf("exchange token: %w", err)
+		return xerrors.Errorf("refresh token: %w", err)
 	}
-	a.sessionToken.Store(&sessionToken)
 
 	// ConnectRPC returns the dRPC connection we use for the Agent and Tailnet v2+ APIs
 	aAPI, tAPI, err := a.client.ConnectRPC26(a.hardCtx)
@@ -1089,7 +1127,7 @@ func (a *agent) handleManifest(manifestOK *checkpoint) func(ctx context.Context,
 		if err != nil {
 			return xerrors.Errorf("fetch metadata: %w", err)
 		}
-		a.logger.Info(ctx, "fetched manifest", slog.F("manifest", mp))
+		a.logger.Info(ctx, "fetched manifest")
 		manifest, err := agentsdk.ManifestFromProto(mp)
 		if err != nil {
 			a.logger.Critical(ctx, "failed to convert manifest", slog.F("manifest", mp), slog.Error(err))
@@ -1163,7 +1201,7 @@ func (a *agent) handleManifest(manifestOK *checkpoint) func(ctx context.Context,
 				scripts             = manifest.Scripts
 				devcontainerScripts map[uuid.UUID]codersdk.WorkspaceAgentScript
 			)
-			if a.containerAPI != nil {
+			if a.devcontainers {
 				// Init the container API with the manifest and client so that
 				// we can start accepting requests. The final start of the API
 				// happens after the startup scripts have been executed to
@@ -1171,7 +1209,7 @@ func (a *agent) handleManifest(manifestOK *checkpoint) func(ctx context.Context,
 				// return existing devcontainers but actual container detection
 				// and creation will be deferred.
 				a.containerAPI.Init(
-					agentcontainers.WithManifestInfo(manifest.OwnerName, manifest.WorkspaceName, manifest.AgentName),
+					agentcontainers.WithManifestInfo(manifest.OwnerName, manifest.WorkspaceName, manifest.AgentName, manifest.Directory),
 					agentcontainers.WithDevcontainers(manifest.Devcontainers, manifest.Scripts),
 					agentcontainers.WithSubAgentClient(agentcontainers.NewSubAgentClientFromAPI(a.logger, aAPI)),
 				)
@@ -1198,7 +1236,7 @@ func (a *agent) handleManifest(manifestOK *checkpoint) func(ctx context.Context,
 				// autostarted devcontainer will be included in this time.
 				err := a.scriptRunner.Execute(a.gracefulCtx, agentscripts.ExecuteStartScripts)
 
-				if a.containerAPI != nil {
+				if a.devcontainers {
 					// Start the container API after the startup scripts have
 					// been executed to ensure that the required tools can be
 					// installed.
@@ -1362,7 +1400,7 @@ func (a *agent) updateCommandEnv(current []string) (updated []string, err error)
 		"CODER_WORKSPACE_OWNER_NAME": manifest.OwnerName,
 
 		// Specific Coder subcommands require the agent token exposed!
-		"CODER_AGENT_TOKEN": *a.sessionToken.Load(),
+		"CODER_AGENT_TOKEN": a.client.GetSessionToken(),
 
 		// Git on Windows resolves with UNIX-style paths.
 		// If using backslashes, it's unable to find the executable.
@@ -1538,8 +1576,8 @@ func (a *agent) createTailnet(
 				break
 			}
 			clog := a.logger.Named("speedtest").With(
-				slog.F("remote", conn.RemoteAddr().String()),
-				slog.F("local", conn.LocalAddr().String()))
+				slog.F("remote", conn.RemoteAddr()),
+				slog.F("local", conn.LocalAddr()))
 			clog.Info(ctx, "accepted conn")
 			wg.Add(1)
 			closed := make(chan struct{})
@@ -1922,6 +1960,7 @@ func (a *agent) Close() error {
 			lifecycleState = codersdk.WorkspaceAgentLifecycleShutdownError
 		}
 	}
+
 	a.setLifecycle(lifecycleState)
 
 	err = a.scriptRunner.Close()
@@ -1929,12 +1968,16 @@ func (a *agent) Close() error {
 		a.logger.Error(a.hardCtx, "script runner close", slog.Error(err))
 	}
 
-	if a.containerAPI != nil {
-		if err := a.containerAPI.Close(); err != nil {
-			a.logger.Error(a.hardCtx, "container API close", slog.Error(err))
+	if a.socketServer != nil {
+		if err := a.socketServer.Close(); err != nil {
+			a.logger.Error(a.hardCtx, "socket server close", slog.Error(err))
 		}
 	}
 
+	if err := a.containerAPI.Close(); err != nil {
+		a.logger.Error(a.hardCtx, "container API close", slog.Error(err))
+	}
+
 	// Wait for the graceful shutdown to complete, but don't wait forever so
 	// that we don't break user expectations.
 	go func() {

agent/agent_internal_test.go

@@ -0,0 +1,45 @@
+package agent
+
+import (
+	"testing"
+
+	"github.com/google/uuid"
+	"github.com/stretchr/testify/require"
+
+	"cdr.dev/slog"
+	"cdr.dev/slog/sloggers/slogtest"
+
+	"github.com/coder/coder/v2/agent/proto"
+	"github.com/coder/coder/v2/testutil"
+)
+
+// TestReportConnectionEmpty tests that reportConnection() doesn't choke if given an empty IP string, which is what we
+// send if we cannot get the remote address.
+func TestReportConnectionEmpty(t *testing.T) {
+	t.Parallel()
+	connID := uuid.UUID{1}
+	logger := slogtest.Make(t, &slogtest.Options{IgnoreErrors: true}).Leveled(slog.LevelDebug)
+	ctx := testutil.Context(t, testutil.WaitShort)
+
+	uut := &agent{
+		hardCtx: ctx,
+		logger:  logger,
+	}
+	disconnected := uut.reportConnection(connID, proto.Connection_TYPE_UNSPECIFIED, "")
+
+	require.Len(t, uut.reportConnections, 1)
+	req0 := uut.reportConnections[0]
+	require.Equal(t, proto.Connection_TYPE_UNSPECIFIED, req0.GetConnection().GetType())
+	require.Equal(t, "", req0.GetConnection().Ip)
+	require.Equal(t, connID[:], req0.GetConnection().GetId())
+	require.Equal(t, proto.Connection_CONNECT, req0.GetConnection().GetAction())
+
+	disconnected(0, "because")
+	require.Len(t, uut.reportConnections, 2)
+	req1 := uut.reportConnections[1]
+	require.Equal(t, proto.Connection_TYPE_UNSPECIFIED, req1.GetConnection().GetType())
+	require.Equal(t, "", req1.GetConnection().Ip)
+	require.Equal(t, connID[:], req1.GetConnection().GetId())
+	require.Equal(t, proto.Connection_DISCONNECT, req1.GetConnection().GetAction())
+	require.Equal(t, "because", req1.GetConnection().GetReason())
+}

agent/agent_test.go

@@ -22,7 +22,6 @@ import (
 	"slices"
 	"strconv"
 	"strings"
-	"sync/atomic"
 	"testing"
 	"time"
 
@@ -456,8 +455,6 @@ func TestAgent_GitSSH(t *testing.T) {
 
 func TestAgent_SessionTTYShell(t *testing.T) {
 	t.Parallel()
-	ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
-	t.Cleanup(cancel)
 	if runtime.GOOS == "windows" {
 		// This might be our implementation, or ConPTY itself.
 		// It's difficult to find extensive tests for it, so
@@ -468,6 +465,7 @@ func TestAgent_SessionTTYShell(t *testing.T) {
 	for _, port := range sshPorts {
 		t.Run(fmt.Sprintf("(%d)", port), func(t *testing.T) {
 			t.Parallel()
+			ctx := testutil.Context(t, testutil.WaitShort)
 
 			session := setupSSHSessionOnPort(t, agentsdk.Manifest{}, codersdk.ServiceBannerConfig{}, nil, port)
 			command := "sh"
@@ -1809,11 +1807,12 @@ func TestAgent_ReconnectingPTY(t *testing.T) {
 
 			//nolint:dogsled
 			conn, agentClient, _, _, _ := setupAgent(t, agentsdk.Manifest{}, 0)
+			idConnectionReport := uuid.New()
 			id := uuid.New()
 
 			// Test that the connection is reported. This must be tested in the
 			// first connection because we care about verifying all of these.
-			netConn0, err := conn.ReconnectingPTY(ctx, id, 80, 80, "bash --norc")
+			netConn0, err := conn.ReconnectingPTY(ctx, idConnectionReport, 80, 80, "bash --norc")
 			require.NoError(t, err)
 			_ = netConn0.Close()
 			assertConnectionReport(t, agentClient, proto.Connection_RECONNECTING_PTY, 0, "")
@@ -2029,7 +2028,8 @@ func runSubAgentMain() int {
 	ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
 	defer cancel()
 	req = req.WithContext(ctx)
-	resp, err := http.DefaultClient.Do(req)
+	client := &http.Client{}
+	resp, err := client.Do(req)
 	if err != nil {
 		_, _ = fmt.Fprintf(os.Stderr, "agent connection failed: %v\n", err)
 		return 11
@@ -2130,7 +2130,7 @@ func TestAgent_DevcontainerAutostart(t *testing.T) {
 		"name": "mywork",
 		"image": "ubuntu:latest",
 		"cmd": ["sleep", "infinity"],
-		"runArgs": ["--network=host"]
+		"runArgs": ["--network=host", "--label=`+agentcontainers.DevcontainerIsTestRunLabel+`=true"]
     }`), 0o600)
 	require.NoError(t, err, "write devcontainer.json")
 
@@ -2167,6 +2167,7 @@ func TestAgent_DevcontainerAutostart(t *testing.T) {
 			// Only match this specific dev container.
 			agentcontainers.WithClock(mClock),
 			agentcontainers.WithContainerLabelIncludeFilter("devcontainer.local_folder", tempWorkspaceFolder),
+			agentcontainers.WithContainerLabelIncludeFilter(agentcontainers.DevcontainerIsTestRunLabel, "true"),
 			agentcontainers.WithSubAgentURL(srv.URL),
 			// The agent will copy "itself", but in the case of this test, the
 			// agent is actually this test binary. So we'll tell the test binary
@@ -2288,7 +2289,8 @@ func TestAgent_DevcontainerRecreate(t *testing.T) {
 	err = os.WriteFile(devcontainerFile, []byte(`{
         "name": "mywork",
         "image": "busybox:latest",
-        "cmd": ["sleep", "infinity"]
+        "cmd": ["sleep", "infinity"],
+		"runArgs": ["--label=`+agentcontainers.DevcontainerIsTestRunLabel+`=true"]
     }`), 0o600)
 	require.NoError(t, err, "write devcontainer.json")
 
@@ -2315,6 +2317,7 @@ func TestAgent_DevcontainerRecreate(t *testing.T) {
 		o.Devcontainers = true
 		o.DevcontainerAPIOptions = append(o.DevcontainerAPIOptions,
 			agentcontainers.WithContainerLabelIncludeFilter("devcontainer.local_folder", workspaceFolder),
+			agentcontainers.WithContainerLabelIncludeFilter(agentcontainers.DevcontainerIsTestRunLabel, "true"),
 		)
 	})
 
@@ -2369,7 +2372,7 @@ func TestAgent_DevcontainerRecreate(t *testing.T) {
 	// devcontainer, we do it in a goroutine so we can process logs
 	// concurrently.
 	go func(container codersdk.WorkspaceAgentContainer) {
-		_, err := conn.RecreateDevcontainer(ctx, container.ID)
+		_, err := conn.RecreateDevcontainer(ctx, devcontainerID.String())
 		assert.NoError(t, err, "recreate devcontainer should succeed")
 	}(container)
 
@@ -2438,7 +2441,8 @@ func TestAgent_DevcontainersDisabledForSubAgent(t *testing.T) {
 
 	// Setup the agent with devcontainers enabled initially.
 	//nolint:dogsled
-	conn, _, _, _, _ := setupAgent(t, manifest, 0, func(*agenttest.Client, *agent.Options) {
+	conn, _, _, _, _ := setupAgent(t, manifest, 0, func(_ *agenttest.Client, o *agent.Options) {
+		o.Devcontainers = true
 	})
 
 	// Query the containers API endpoint. This should fail because
@@ -2450,8 +2454,214 @@ func TestAgent_DevcontainersDisabledForSubAgent(t *testing.T) {
 	require.Error(t, err)
 
 	// Verify the error message contains the expected text.
-	require.Contains(t, err.Error(), "The agent dev containers feature is experimental and not enabled by default.")
-	require.Contains(t, err.Error(), "To enable this feature, set CODER_AGENT_DEVCONTAINERS_ENABLE=true in your template.")
+	require.Contains(t, err.Error(), "Dev Container feature not supported.")
+	require.Contains(t, err.Error(), "Dev Container integration inside other Dev Containers is explicitly not supported.")
+}
+
+// TestAgent_DevcontainerPrebuildClaim tests that we correctly handle
+// the claiming process for running devcontainers.
+//
+// You can run it manually as follows:
+//
+// CODER_TEST_USE_DOCKER=1 go test -count=1 ./agent -run TestAgent_DevcontainerPrebuildClaim
+//
+//nolint:paralleltest // This test sets an environment variable.
+func TestAgent_DevcontainerPrebuildClaim(t *testing.T) {
+	if os.Getenv("CODER_TEST_USE_DOCKER") != "1" {
+		t.Skip("Set CODER_TEST_USE_DOCKER=1 to run this test")
+	}
+	if _, err := exec.LookPath("devcontainer"); err != nil {
+		t.Skip("This test requires the devcontainer CLI: npm install -g @devcontainers/cli")
+	}
+
+	pool, err := dockertest.NewPool("")
+	require.NoError(t, err, "Could not connect to docker")
+
+	var (
+		ctx = testutil.Context(t, testutil.WaitShort)
+
+		devcontainerID          = uuid.New()
+		devcontainerLogSourceID = uuid.New()
+
+		workspaceFolder    = filepath.Join(t.TempDir(), "project")
+		devcontainerPath   = filepath.Join(workspaceFolder, ".devcontainer")
+		devcontainerConfig = filepath.Join(devcontainerPath, "devcontainer.json")
+	)
+
+	// Given: A devcontainer project.
+	t.Logf("Workspace folder: %s", workspaceFolder)
+
+	err = os.MkdirAll(devcontainerPath, 0o755)
+	require.NoError(t, err, "create dev container directory")
+
+	// Given: This devcontainer project specifies an app that uses the owner name and workspace name.
+	err = os.WriteFile(devcontainerConfig, []byte(`{
+	    "name": "project",
+		"image": "busybox:latest",
+	    "cmd": ["sleep", "infinity"],
+		"runArgs": ["--label=`+agentcontainers.DevcontainerIsTestRunLabel+`=true"],
+		"customizations": {
+		 	"coder": {
+				"apps": [{
+					"slug": "zed",
+					"url": "zed://ssh/${localEnv:CODER_WORKSPACE_AGENT_NAME}.${localEnv:CODER_WORKSPACE_NAME}.${localEnv:CODER_WORKSPACE_OWNER_NAME}.coder${containerWorkspaceFolder}"
+				}]
+			}
+		}
+	}`), 0o600)
+	require.NoError(t, err, "write devcontainer config")
+
+	// Given: A manifest with a prebuild username and workspace name.
+	manifest := agentsdk.Manifest{
+		OwnerName:     "prebuilds",
+		WorkspaceName: "prebuilds-xyz-123",
+
+		Devcontainers: []codersdk.WorkspaceAgentDevcontainer{
+			{ID: devcontainerID, Name: "test", WorkspaceFolder: workspaceFolder},
+		},
+		Scripts: []codersdk.WorkspaceAgentScript{
+			{ID: devcontainerID, LogSourceID: devcontainerLogSourceID},
+		},
+	}
+
+	// When: We create an agent with devcontainers enabled.
+	//nolint:dogsled
+	conn, client, _, _, _ := setupAgent(t, manifest, 0, func(_ *agenttest.Client, o *agent.Options) {
+		o.Devcontainers = true
+		o.DevcontainerAPIOptions = append(o.DevcontainerAPIOptions,
+			agentcontainers.WithContainerLabelIncludeFilter(agentcontainers.DevcontainerLocalFolderLabel, workspaceFolder),
+			agentcontainers.WithContainerLabelIncludeFilter(agentcontainers.DevcontainerIsTestRunLabel, "true"),
+		)
+	})
+
+	testutil.Eventually(ctx, t, func(ctx context.Context) bool {
+		return slices.Contains(client.GetLifecycleStates(), codersdk.WorkspaceAgentLifecycleReady)
+	}, testutil.IntervalMedium, "agent not ready")
+
+	var dcPrebuild codersdk.WorkspaceAgentDevcontainer
+	testutil.Eventually(ctx, t, func(ctx context.Context) bool {
+		resp, err := conn.ListContainers(ctx)
+		require.NoError(t, err)
+
+		for _, dc := range resp.Devcontainers {
+			if dc.Container == nil {
+				continue
+			}
+
+			v, ok := dc.Container.Labels[agentcontainers.DevcontainerLocalFolderLabel]
+			if ok && v == workspaceFolder {
+				dcPrebuild = dc
+				return true
+			}
+		}
+
+		return false
+	}, testutil.IntervalMedium, "devcontainer not found")
+	defer func() {
+		pool.Client.RemoveContainer(docker.RemoveContainerOptions{
+			ID:            dcPrebuild.Container.ID,
+			RemoveVolumes: true,
+			Force:         true,
+		})
+	}()
+
+	// Then: We expect a sub agent to have been created.
+	subAgents := client.GetSubAgents()
+	require.Len(t, subAgents, 1)
+
+	subAgent := subAgents[0]
+	subAgentID, err := uuid.FromBytes(subAgent.GetId())
+	require.NoError(t, err)
+
+	// And: We expect there to be 1 app.
+	subAgentApps, err := client.GetSubAgentApps(subAgentID)
+	require.NoError(t, err)
+	require.Len(t, subAgentApps, 1)
+
+	// And: This app should contain the prebuild workspace name and owner name.
+	subAgentApp := subAgentApps[0]
+	require.Equal(t, "zed://ssh/project.prebuilds-xyz-123.prebuilds.coder/workspaces/project", subAgentApp.GetUrl())
+
+	// Given: We close the client and connection
+	client.Close()
+	conn.Close()
+
+	// Given: A new manifest with a regular user owner name and workspace name.
+	manifest = agentsdk.Manifest{
+		OwnerName:     "user",
+		WorkspaceName: "user-workspace",
+
+		Devcontainers: []codersdk.WorkspaceAgentDevcontainer{
+			{ID: devcontainerID, Name: "test", WorkspaceFolder: workspaceFolder},
+		},
+		Scripts: []codersdk.WorkspaceAgentScript{
+			{ID: devcontainerID, LogSourceID: devcontainerLogSourceID},
+		},
+	}
+
+	// When: We create an agent with devcontainers enabled.
+	//nolint:dogsled
+	conn, client, _, _, _ = setupAgent(t, manifest, 0, func(_ *agenttest.Client, o *agent.Options) {
+		o.Devcontainers = true
+		o.DevcontainerAPIOptions = append(o.DevcontainerAPIOptions,
+			agentcontainers.WithContainerLabelIncludeFilter(agentcontainers.DevcontainerLocalFolderLabel, workspaceFolder),
+			agentcontainers.WithContainerLabelIncludeFilter(agentcontainers.DevcontainerIsTestRunLabel, "true"),
+		)
+	})
+
+	testutil.Eventually(ctx, t, func(ctx context.Context) bool {
+		return slices.Contains(client.GetLifecycleStates(), codersdk.WorkspaceAgentLifecycleReady)
+	}, testutil.IntervalMedium, "agent not ready")
+
+	var dcClaimed codersdk.WorkspaceAgentDevcontainer
+	testutil.Eventually(ctx, t, func(ctx context.Context) bool {
+		resp, err := conn.ListContainers(ctx)
+		require.NoError(t, err)
+
+		for _, dc := range resp.Devcontainers {
+			if dc.Container == nil {
+				continue
+			}
+
+			v, ok := dc.Container.Labels[agentcontainers.DevcontainerLocalFolderLabel]
+			if ok && v == workspaceFolder {
+				dcClaimed = dc
+				return true
+			}
+		}
+
+		return false
+	}, testutil.IntervalMedium, "devcontainer not found")
+	defer func() {
+		if dcClaimed.Container.ID != dcPrebuild.Container.ID {
+			pool.Client.RemoveContainer(docker.RemoveContainerOptions{
+				ID:            dcClaimed.Container.ID,
+				RemoveVolumes: true,
+				Force:         true,
+			})
+		}
+	}()
+
+	// Then: We expect the claimed devcontainer and prebuild devcontainer
+	// to be using the same underlying container.
+	require.Equal(t, dcPrebuild.Container.ID, dcClaimed.Container.ID)
+
+	// And: We expect there to be a sub agent created.
+	subAgents = client.GetSubAgents()
+	require.Len(t, subAgents, 1)
+
+	subAgent = subAgents[0]
+	subAgentID, err = uuid.FromBytes(subAgent.GetId())
+	require.NoError(t, err)
+
+	// And: We expect there to be an app.
+	subAgentApps, err = client.GetSubAgentApps(subAgentID)
+	require.NoError(t, err)
+	require.Len(t, subAgentApps, 1)
+
+	// And: We expect this app to have the user's owner name and workspace name.
+	subAgentApp = subAgentApps[0]
+	require.Equal(t, "zed://ssh/project.user-workspace.user.coder/workspaces/project", subAgentApp.GetUrl())
 }
 
 func TestAgent_Dial(t *testing.T) {
@@ -2459,11 +2669,11 @@ func TestAgent_Dial(t *testing.T) {
 
 	cases := []struct {
 		name  string
-		setup func(t *testing.T) net.Listener
+		setup func(t testing.TB) net.Listener
 	}{
 		{
 			name: "TCP",
-			setup: func(t *testing.T) net.Listener {
+			setup: func(t testing.TB) net.Listener {
 				l, err := net.Listen("tcp", "127.0.0.1:0")
 				require.NoError(t, err, "create TCP listener")
 				return l
@@ -2471,7 +2681,7 @@ func TestAgent_Dial(t *testing.T) {
 		},
 		{
 			name: "UDP",
-			setup: func(t *testing.T) net.Listener {
+			setup: func(t testing.TB) net.Listener {
 				addr := net.UDPAddr{
 					IP:   net.ParseIP("127.0.0.1"),
 					Port: 0,
@@ -2489,57 +2699,69 @@ func TestAgent_Dial(t *testing.T) {
 
 			// The purpose of this test is to ensure that a client can dial a
 			// listener in the workspace over tailnet.
-			l := c.setup(t)
-			done := make(chan struct{})
-			defer func() {
-				l.Close()
-				<-done
-			}()
+			//
+			// The OS sometimes drops packets if the system can't keep up with
+			// them. For TCP packets, it's typically fine due to
+			// retransmissions, but for UDP packets, it can fail this test.
+			//
+			// The OS gets involved for the Wireguard traffic (either via DERP
+			// or direct UDP), and also for the traffic between the agent and
+			// the listener in the "workspace".
+			//
+			// To avoid this, we'll retry this test up to 3 times.
+			//nolint:gocritic // This test is flaky due to uncontrollable OS packet drops under heavy load.
+			testutil.RunRetry(t, 3, func(t testing.TB) {
+				ctx := testutil.Context(t, testutil.WaitLong)
 
-			ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
-			defer cancel()
+				l := c.setup(t)
+				done := make(chan struct{})
+				defer func() {
+					l.Close()
+					<-done
+				}()
 
-			go func() {
-				defer close(done)
-				for range 2 {
-					c, err := l.Accept()
-					if assert.NoError(t, err, "accept connection") {
-						testAccept(ctx, t, c)
-						_ = c.Close()
+				go func() {
+					defer close(done)
+					for range 2 {
+						c, err := l.Accept()
+						if assert.NoError(t, err, "accept connection") {
+							testAccept(ctx, t, c)
+							_ = c.Close()
+						}
 					}
+				}()
+
+				agentID := uuid.UUID{0, 0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 7, 8}
+				//nolint:dogsled
+				agentConn, _, _, _, _ := setupAgent(t, agentsdk.Manifest{
+					AgentID: agentID,
+				}, 0)
+				require.True(t, agentConn.AwaitReachable(ctx))
+				conn, err := agentConn.DialContext(ctx, l.Addr().Network(), l.Addr().String())
+				require.NoError(t, err)
+				testDial(ctx, t, conn)
+				err = conn.Close()
+				require.NoError(t, err)
+
+				// also connect via the CoderServicePrefix, to test that we can reach the agent on this
+				// IP. This will be required for CoderVPN.
+				_, rawPort, _ := net.SplitHostPort(l.Addr().String())
+				port, _ := strconv.ParseUint(rawPort, 10, 16)
+				ipp := netip.AddrPortFrom(tailnet.CoderServicePrefix.AddrFromUUID(agentID), uint16(port))
+
+				switch l.Addr().Network() {
+				case "tcp":
+					conn, err = agentConn.TailnetConn().DialContextTCP(ctx, ipp)
+				case "udp":
+					conn, err = agentConn.TailnetConn().DialContextUDP(ctx, ipp)
+				default:
+					t.Fatalf("unknown network: %s", l.Addr().Network())
 				}
-			}()
-
-			agentID := uuid.UUID{0, 0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 7, 8}
-			//nolint:dogsled
-			agentConn, _, _, _, _ := setupAgent(t, agentsdk.Manifest{
-				AgentID: agentID,
-			}, 0)
-			require.True(t, agentConn.AwaitReachable(ctx))
-			conn, err := agentConn.DialContext(ctx, l.Addr().Network(), l.Addr().String())
-			require.NoError(t, err)
-			testDial(ctx, t, conn)
-			err = conn.Close()
-			require.NoError(t, err)
-
-			// also connect via the CoderServicePrefix, to test that we can reach the agent on this
-			// IP. This will be required for CoderVPN.
-			_, rawPort, _ := net.SplitHostPort(l.Addr().String())
-			port, _ := strconv.ParseUint(rawPort, 10, 16)
-			ipp := netip.AddrPortFrom(tailnet.CoderServicePrefix.AddrFromUUID(agentID), uint16(port))
-
-			switch l.Addr().Network() {
-			case "tcp":
-				conn, err = agentConn.Conn.DialContextTCP(ctx, ipp)
-			case "udp":
-				conn, err = agentConn.Conn.DialContextUDP(ctx, ipp)
-			default:
-				t.Fatalf("unknown network: %s", l.Addr().Network())
-			}
-			require.NoError(t, err)
-			testDial(ctx, t, conn)
-			err = conn.Close()
-			require.NoError(t, err)
+				require.NoError(t, err)
+				testDial(ctx, t, conn)
+				err = conn.Close()
+				require.NoError(t, err)
+			})
 		})
 	}
 }
@@ -2590,7 +2812,7 @@ func TestAgent_UpdatedDERP(t *testing.T) {
 	})
 
 	// Setup a client connection.
-	newClientConn := func(derpMap *tailcfg.DERPMap, name string) *workspacesdk.AgentConn {
+	newClientConn := func(derpMap *tailcfg.DERPMap, name string) workspacesdk.AgentConn {
 		conn, err := tailnet.NewConn(&tailnet.Options{
 			Addresses: []netip.Prefix{tailnet.TailscaleServicePrefix.RandomPrefix()},
 			DERPMap:   derpMap,
@@ -2670,13 +2892,13 @@ func TestAgent_UpdatedDERP(t *testing.T) {
 
 	// Connect from a second client and make sure it uses the new DERP map.
 	conn2 := newClientConn(newDerpMap, "client2")
-	require.Equal(t, []int{2}, conn2.DERPMap().RegionIDs())
+	require.Equal(t, []int{2}, conn2.TailnetConn().DERPMap().RegionIDs())
 	t.Log("conn2 got the new DERPMap")
 
 	// If the first client gets a DERP map update, it should be able to
 	// reconnect just fine.
-	conn1.SetDERPMap(newDerpMap)
-	require.Equal(t, []int{2}, conn1.DERPMap().RegionIDs())
+	conn1.TailnetConn().SetDERPMap(newDerpMap)
+	require.Equal(t, []int{2}, conn1.TailnetConn().DERPMap().RegionIDs())
 	t.Log("set the new DERPMap on conn1")
 	ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
 	defer cancel()
@@ -2705,11 +2927,11 @@ func TestAgent_Speedtest(t *testing.T) {
 
 func TestAgent_Reconnect(t *testing.T) {
 	t.Parallel()
+	ctx := testutil.Context(t, testutil.WaitShort)
 	logger := testutil.Logger(t)
 	// After the agent is disconnected from a coordinator, it's supposed
 	// to reconnect!
-	coordinator := tailnet.NewCoordinator(logger)
-	defer coordinator.Close()
+	fCoordinator := tailnettest.NewFakeCoordinator()
 
 	agentID := uuid.New()
 	statsCh := make(chan *proto.Stats, 50)
@@ -2721,27 +2943,24 @@ func TestAgent_Reconnect(t *testing.T) {
 			DERPMap: derpMap,
 		},
 		statsCh,
-		coordinator,
+		fCoordinator,
 	)
 	defer client.Close()
-	initialized := atomic.Int32{}
+
 	closer := agent.New(agent.Options{
-		ExchangeToken: func(ctx context.Context) (string, error) {
-			initialized.Add(1)
-			return "", nil
-		},
 		Client: client,
 		Logger: logger.Named("agent"),
 	})
 	defer closer.Close()
 
-	require.Eventually(t, func() bool {
-		return coordinator.Node(agentID) != nil
-	}, testutil.WaitShort, testutil.IntervalFast)
-	client.LastWorkspaceAgent()
-	require.Eventually(t, func() bool {
-		return initialized.Load() == 2
-	}, testutil.WaitShort, testutil.IntervalFast)
+	call1 := testutil.RequireReceive(ctx, t, fCoordinator.CoordinateCalls)
+	require.Equal(t, client.GetNumRefreshTokenCalls(), 1)
+	close(call1.Resps) // hang up
+	// expect reconnect
+	testutil.RequireReceive(ctx, t, fCoordinator.CoordinateCalls)
+	// Check that the agent refreshes the token when it reconnects.
+	require.Equal(t, client.GetNumRefreshTokenCalls(), 2)
+	closer.Close()
 }
 
 func TestAgent_WriteVSCodeConfigs(t *testing.T) {
@@ -2763,9 +2982,6 @@ func TestAgent_WriteVSCodeConfigs(t *testing.T) {
 	defer client.Close()
 	filesystem := afero.NewMemMapFs()
 	closer := agent.New(agent.Options{
-		ExchangeToken: func(ctx context.Context) (string, error) {
-			return "", nil
-		},
 		Client:     client,
 		Logger:     logger.Named("agent"),
 		Filesystem: filesystem,
@@ -2794,9 +3010,6 @@ func TestAgent_DebugServer(t *testing.T) {
 	conn, _, _, _, agnt := setupAgent(t, agentsdk.Manifest{
 		DERPMap: derpMap,
 	}, 0, func(c *agenttest.Client, o *agent.Options) {
-		o.ExchangeToken = func(context.Context) (string, error) {
-			return "token", nil
-		}
 		o.LogDir = logDir
 	})
 
@@ -3042,8 +3255,8 @@ func setupSSHSessionOnPort(
 	return session
 }
 
-func setupAgent(t *testing.T, metadata agentsdk.Manifest, ptyTimeout time.Duration, opts ...func(*agenttest.Client, *agent.Options)) (
-	*workspacesdk.AgentConn,
+func setupAgent(t testing.TB, metadata agentsdk.Manifest, ptyTimeout time.Duration, opts ...func(*agenttest.Client, *agent.Options)) (
+	workspacesdk.AgentConn,
 	*agenttest.Client,
 	<-chan *proto.Stats,
 	afero.Fs,
@@ -3140,7 +3353,7 @@ func setupAgent(t *testing.T, metadata agentsdk.Manifest, ptyTimeout time.Durati
 
 var dialTestPayload = []byte("dean-was-here123")
 
-func testDial(ctx context.Context, t *testing.T, c net.Conn) {
+func testDial(ctx context.Context, t testing.TB, c net.Conn) {
 	t.Helper()
 
 	if deadline, ok := ctx.Deadline(); ok {
@@ -3156,7 +3369,7 @@ func testDial(ctx context.Context, t *testing.T, c net.Conn) {
 	assertReadPayload(t, c, dialTestPayload)
 }
 
-func testAccept(ctx context.Context, t *testing.T, c net.Conn) {
+func testAccept(ctx context.Context, t testing.TB, c net.Conn) {
 	t.Helper()
 	defer c.Close()
 
@@ -3173,7 +3386,7 @@ func testAccept(ctx context.Context, t *testing.T, c net.Conn) {
 	assertWritePayload(t, c, dialTestPayload)
 }
 
-func assertReadPayload(t *testing.T, r io.Reader, payload []byte) {
+func assertReadPayload(t testing.TB, r io.Reader, payload []byte) {
 	t.Helper()
 	b := make([]byte, len(payload)+16)
 	n, err := r.Read(b)
@@ -3182,11 +3395,11 @@ func assertReadPayload(t *testing.T, r io.Reader, payload []byte) {
 	assert.Equal(t, payload, b[:n])
 }
 
-func assertWritePayload(t *testing.T, w io.Writer, payload []byte) {
+func assertWritePayload(t testing.TB, w io.Writer, payload []byte) {
 	t.Helper()
 	n, err := w.Write(payload)
 	assert.NoError(t, err, "write payload")
-	assert.Equal(t, len(payload), n, "payload length does not match")
+	assert.Equal(t, len(payload), n, "written payload length does not match")
 }
 
 func testSessionOutput(t *testing.T, session *ssh.Session, expected, unexpected []string, expectedRe *regexp.Regexp) {
@@ -3264,16 +3477,31 @@ func TestAgent_Metrics_SSH(t *testing.T) {
 	err = session.Shell()
 	require.NoError(t, err)
 
-	expected := []*proto.Stats_Metric{
+	expected := []struct {
+		Name    string
+		Type    proto.Stats_Metric_Type
+		CheckFn func(float64) error
+		Labels  []*proto.Stats_Metric_Label
+	}{
 		{
-			Name:  "agent_reconnecting_pty_connections_total",
-			Type:  proto.Stats_Metric_COUNTER,
-			Value: 0,
+			Name: "agent_reconnecting_pty_connections_total",
+			Type: proto.Stats_Metric_COUNTER,
+			CheckFn: func(v float64) error {
+				if v == 0 {
+					return nil
+				}
+				return xerrors.Errorf("expected 0, got %f", v)
+			},
 		},
 		{
-			Name:  "agent_sessions_total",
-			Type:  proto.Stats_Metric_COUNTER,
-			Value: 1,
+			Name: "agent_sessions_total",
+			Type: proto.Stats_Metric_COUNTER,
+			CheckFn: func(v float64) error {
+				if v == 1 {
+					return nil
+				}
+				return xerrors.Errorf("expected 1, got %f", v)
+			},
 			Labels: []*proto.Stats_Metric_Label{
 				{
 					Name:  "magic_type",
@@ -3286,24 +3514,44 @@ func TestAgent_Metrics_SSH(t *testing.T) {
 			},
 		},
 		{
-			Name:  "agent_ssh_server_failed_connections_total",
-			Type:  proto.Stats_Metric_COUNTER,
-			Value: 0,
+			Name: "agent_ssh_server_failed_connections_total",
+			Type: proto.Stats_Metric_COUNTER,
+			CheckFn: func(v float64) error {
+				if v == 0 {
+					return nil
+				}
+				return xerrors.Errorf("expected 0, got %f", v)
+			},
 		},
 		{
-			Name:  "agent_ssh_server_sftp_connections_total",
-			Type:  proto.Stats_Metric_COUNTER,
-			Value: 0,
+			Name: "agent_ssh_server_sftp_connections_total",
+			Type: proto.Stats_Metric_COUNTER,
+			CheckFn: func(v float64) error {
+				if v == 0 {
+					return nil
+				}
+				return xerrors.Errorf("expected 0, got %f", v)
+			},
 		},
 		{
-			Name:  "agent_ssh_server_sftp_server_errors_total",
-			Type:  proto.Stats_Metric_COUNTER,
-			Value: 0,
+			Name: "agent_ssh_server_sftp_server_errors_total",
+			Type: proto.Stats_Metric_COUNTER,
+			CheckFn: func(v float64) error {
+				if v == 0 {
+					return nil
+				}
+				return xerrors.Errorf("expected 0, got %f", v)
+			},
 		},
 		{
-			Name:  "coderd_agentstats_currently_reachable_peers",
-			Type:  proto.Stats_Metric_GAUGE,
-			Value: 0,
+			Name: "coderd_agentstats_currently_reachable_peers",
+			Type: proto.Stats_Metric_GAUGE,
+			CheckFn: func(float64) error {
+				// We can't reliably ping a peer here, and networking is out of
+				// scope of this test, so we just test that the metric exists
+				// with the correct labels.
+				return nil
+			},
 			Labels: []*proto.Stats_Metric_Label{
 				{
 					Name:  "connection_type",
@@ -3312,9 +3560,11 @@ func TestAgent_Metrics_SSH(t *testing.T) {
 			},
 		},
 		{
-			Name:  "coderd_agentstats_currently_reachable_peers",
-			Type:  proto.Stats_Metric_GAUGE,
-			Value: 1,
+			Name: "coderd_agentstats_currently_reachable_peers",
+			Type: proto.Stats_Metric_GAUGE,
+			CheckFn: func(float64) error {
+				return nil
+			},
 			Labels: []*proto.Stats_Metric_Label{
 				{
 					Name:  "connection_type",
@@ -3323,9 +3573,20 @@ func TestAgent_Metrics_SSH(t *testing.T) {
 			},
 		},
 		{
-			Name:  "coderd_agentstats_startup_script_seconds",
-			Type:  proto.Stats_Metric_GAUGE,
-			Value: 1,
+			Name: "coderd_agentstats_startup_script_seconds",
+			Type: proto.Stats_Metric_GAUGE,
+			CheckFn: func(f float64) error {
+				if f >= 0 {
+					return nil
+				}
+				return xerrors.Errorf("expected >= 0, got %f", f)
+			},
+			Labels: []*proto.Stats_Metric_Label{
+				{
+					Name:  "success",
+					Value: "true",
+				},
+			},
 		},
 	}
 
@@ -3347,11 +3608,10 @@ func TestAgent_Metrics_SSH(t *testing.T) {
 		for _, m := range mf.GetMetric() {
 			assert.Equal(t, expected[i].Name, mf.GetName())
 			assert.Equal(t, expected[i].Type.String(), mf.GetType().String())
-			// Value is max expected
 			if expected[i].Type == proto.Stats_Metric_GAUGE {
-				assert.GreaterOrEqualf(t, expected[i].Value, m.GetGauge().GetValue(), "expected %s to be greater than or equal to %f, got %f", expected[i].Name, expected[i].Value, m.GetGauge().GetValue())
+				assert.NoError(t, expected[i].CheckFn(m.GetGauge().GetValue()), "check fn for %s failed", expected[i].Name)
 			} else if expected[i].Type == proto.Stats_Metric_COUNTER {
-				assert.GreaterOrEqualf(t, expected[i].Value, m.GetCounter().GetValue(), "expected %s to be greater than or equal to %f, got %f", expected[i].Name, expected[i].Value, m.GetCounter().GetValue())
+				assert.NoError(t, expected[i].CheckFn(m.GetCounter().GetValue()), "check fn for %s failed", expected[i].Name)
 			}
 			for j, lbl := range expected[i].Labels {
 				assert.Equal(t, m.GetLabel()[j], &promgo.LabelPair{

agent/agentcontainers/api.go

@@ -2,8 +2,11 @@ package agentcontainers
 
 import (
 	"context"
+	"encoding/json"
 	"errors"
 	"fmt"
+	"io/fs"
+	"maps"
 	"net/http"
 	"os"
 	"path"
@@ -18,10 +21,13 @@ import (
 
 	"github.com/fsnotify/fsnotify"
 	"github.com/go-chi/chi/v5"
+	"github.com/go-git/go-git/v5/plumbing/format/gitignore"
 	"github.com/google/uuid"
+	"github.com/spf13/afero"
 	"golang.org/x/xerrors"
 
 	"cdr.dev/slog"
+	"github.com/coder/coder/v2/agent/agentcontainers/ignore"
 	"github.com/coder/coder/v2/agent/agentcontainers/watcher"
 	"github.com/coder/coder/v2/agent/agentexec"
 	"github.com/coder/coder/v2/agent/usershell"
@@ -30,6 +36,7 @@ import (
 	"github.com/coder/coder/v2/codersdk/agentsdk"
 	"github.com/coder/coder/v2/provisioner"
 	"github.com/coder/quartz"
+	"github.com/coder/websocket"
 )
 
 const (
@@ -53,10 +60,12 @@ type API struct {
 	cancel                      context.CancelFunc
 	watcherDone                 chan struct{}
 	updaterDone                 chan struct{}
+	discoverDone                chan struct{}
 	updateTrigger               chan chan error // Channel to trigger manual refresh.
 	updateInterval              time.Duration   // Interval for periodic container updates.
 	logger                      slog.Logger
 	watcher                     watcher.Watcher
+	fs                          afero.Fs
 	execer                      agentexec.Execer
 	commandEnv                  CommandEnv
 	ccli                        ContainerCLI
@@ -68,12 +77,17 @@ type API struct {
 	subAgentURL                 string
 	subAgentEnv                 []string
 
-	ownerName     string
-	workspaceName string
-	parentAgent   string
+	projectDiscovery   bool // If we should perform project discovery or not.
+	discoveryAutostart bool // If we should autostart discovered projects.
+
+	ownerName      string
+	workspaceName  string
+	parentAgent    string
+	agentDirectory string
 
 	mu                       sync.RWMutex  // Protects the following fields.
 	initDone                 chan struct{} // Closed by Init.
+	updateChans              []chan struct{}
 	closed                   bool
 	containers               codersdk.WorkspaceAgentListContainersResponse  // Output from the last list operation.
 	containersErr            error                                          // Error from the last list operation.
@@ -130,7 +144,9 @@ func WithCommandEnv(ce CommandEnv) Option {
 					strings.HasPrefix(s, "CODER_WORKSPACE_AGENT_URL=") ||
 					strings.HasPrefix(s, "CODER_AGENT_TOKEN=") ||
 					strings.HasPrefix(s, "CODER_AGENT_AUTH=") ||
-					strings.HasPrefix(s, "CODER_AGENT_DEVCONTAINERS_ENABLE=")
+					strings.HasPrefix(s, "CODER_AGENT_DEVCONTAINERS_ENABLE=") ||
+					strings.HasPrefix(s, "CODER_AGENT_DEVCONTAINERS_PROJECT_DISCOVERY_ENABLE=") ||
+					strings.HasPrefix(s, "CODER_AGENT_DEVCONTAINERS_DISCOVERY_AUTOSTART_ENABLE=")
 			})
 			return shell, dir, env, nil
 		}
@@ -147,8 +163,8 @@ func WithContainerCLI(ccli ContainerCLI) Option {
 
 // WithContainerLabelIncludeFilter sets a label filter for containers.
 // This option can be given multiple times to filter by multiple labels.
-// The behavior is such that only containers matching one or more of the
-// provided labels will be included.
+// The behavior is such that only containers matching all of the provided
+// labels will be included.
 func WithContainerLabelIncludeFilter(label, value string) Option {
 	return func(api *API) {
 		api.containerLabelIncludeFilter[label] = value
@@ -188,11 +204,12 @@ func WithSubAgentEnv(env ...string) Option {
 
 // WithManifestInfo sets the owner name, and workspace name
 // for the sub-agent.
-func WithManifestInfo(owner, workspace, parentAgent string) Option {
+func WithManifestInfo(owner, workspace, parentAgent, agentDirectory string) Option {
 	return func(api *API) {
 		api.ownerName = owner
 		api.workspaceName = workspace
 		api.parentAgent = parentAgent
+		api.agentDirectory = agentDirectory
 	}
 }
 
@@ -257,6 +274,29 @@ func WithWatcher(w watcher.Watcher) Option {
 	}
 }
 
+// WithFileSystem sets the file system used for discovering projects.
+func WithFileSystem(fileSystem afero.Fs) Option {
+	return func(api *API) {
+		api.fs = fileSystem
+	}
+}
+
+// WithProjectDiscovery sets if the API should attempt to discover
+// projects on the filesystem.
+func WithProjectDiscovery(projectDiscovery bool) Option {
+	return func(api *API) {
+		api.projectDiscovery = projectDiscovery
+	}
+}
+
+// WithDiscoveryAutostart sets if the API should attempt to autostart
+// projects that have been discovered
+func WithDiscoveryAutostart(discoveryAutostart bool) Option {
+	return func(api *API) {
+		api.discoveryAutostart = discoveryAutostart
+	}
+}
+
 // ScriptLogger is an interface for sending devcontainer logs to the
 // controlplane.
 type ScriptLogger interface {
@@ -327,6 +367,9 @@ func NewAPI(logger slog.Logger, options ...Option) *API {
 			api.watcher = watcher.NewNoop()
 		}
 	}
+	if api.fs == nil {
+		api.fs = afero.NewOsFs()
+	}
 	if api.subAgentClient.Load() == nil {
 		var c SubAgentClient = noopSubAgentClient{}
 		api.subAgentClient.Store(&c)
@@ -368,6 +411,12 @@ func (api *API) Start() {
 		return
 	}
 
+	if api.projectDiscovery && api.agentDirectory != "" {
+		api.discoverDone = make(chan struct{})
+
+		go api.discover()
+	}
+
 	api.watcherDone = make(chan struct{})
 	api.updaterDone = make(chan struct{})
 
@@ -375,6 +424,162 @@ func (api *API) Start() {
 	go api.updaterLoop()
 }
 
+func (api *API) discover() {
+	defer close(api.discoverDone)
+	defer api.logger.Debug(api.ctx, "project discovery finished")
+	api.logger.Debug(api.ctx, "project discovery started")
+
+	if err := api.discoverDevcontainerProjects(); err != nil {
+		api.logger.Error(api.ctx, "discovering dev container projects", slog.Error(err))
+	}
+
+	if err := api.RefreshContainers(api.ctx); err != nil {
+		api.logger.Error(api.ctx, "refreshing containers after discovery", slog.Error(err))
+	}
+}
+
+func (api *API) discoverDevcontainerProjects() error {
+	isGitProject, err := afero.DirExists(api.fs, filepath.Join(api.agentDirectory, ".git"))
+	if err != nil {
+		return xerrors.Errorf(".git dir exists: %w", err)
+	}
+
+	// If the agent directory is a git project, we'll search
+	// the project for any `.devcontainer/devcontainer.json`
+	// files.
+	if isGitProject {
+		return api.discoverDevcontainersInProject(api.agentDirectory)
+	}
+
+	// The agent directory is _not_ a git project, so we'll
+	// search the top level of the agent directory for any
+	// git projects, and search those.
+	entries, err := afero.ReadDir(api.fs, api.agentDirectory)
+	if err != nil {
+		return xerrors.Errorf("read agent directory: %w", err)
+	}
+
+	for _, entry := range entries {
+		if !entry.IsDir() {
+			continue
+		}
+
+		isGitProject, err = afero.DirExists(api.fs, filepath.Join(api.agentDirectory, entry.Name(), ".git"))
+		if err != nil {
+			return xerrors.Errorf(".git dir exists: %w", err)
+		}
+
+		// If this directory is a git project, we'll search
+		// it for any `.devcontainer/devcontainer.json` files.
+		if isGitProject {
+			if err := api.discoverDevcontainersInProject(filepath.Join(api.agentDirectory, entry.Name())); err != nil {
+				return err
+			}
+		}
+	}
+
+	return nil
+}
+
+func (api *API) discoverDevcontainersInProject(projectPath string) error {
+	logger := api.logger.
+		Named("project-discovery").
+		With(slog.F("project_path", projectPath))
+
+	globalPatterns, err := ignore.LoadGlobalPatterns(api.fs)
+	if err != nil {
+		return xerrors.Errorf("read global git ignore patterns: %w", err)
+	}
+
+	patterns, err := ignore.ReadPatterns(api.ctx, logger, api.fs, projectPath)
+	if err != nil {
+		return xerrors.Errorf("read git ignore patterns: %w", err)
+	}
+
+	matcher := gitignore.NewMatcher(append(globalPatterns, patterns...))
+
+	devcontainerConfigPaths := []string{
+		"/.devcontainer/devcontainer.json",
+		"/.devcontainer.json",
+	}
+
+	return afero.Walk(api.fs, projectPath, func(path string, info fs.FileInfo, err error) error {
+		if err != nil {
+			logger.Error(api.ctx, "encountered error while walking for dev container projects",
+				slog.F("path", path),
+				slog.Error(err))
+			return nil
+		}
+
+		pathParts := ignore.FilePathToParts(path)
+
+		// We know that a directory entry cannot be a `devcontainer.json` file, so we
+		// always skip processing directories. If the directory happens to be ignored
+		// by git then we'll make sure to ignore all of the children of that directory.
+		if info.IsDir() {
+			if matcher.Match(pathParts, true) {
+				return fs.SkipDir
+			}
+
+			return nil
+		}
+
+		if matcher.Match(pathParts, false) {
+			return nil
+		}
+
+		for _, relativeConfigPath := range devcontainerConfigPaths {
+			if !strings.HasSuffix(path, relativeConfigPath) {
+				continue
+			}
+
+			workspaceFolder := strings.TrimSuffix(path, relativeConfigPath)
+
+			logger := logger.With(slog.F("workspace_folder", workspaceFolder))
+			logger.Debug(api.ctx, "discovered dev container project")
+
+			api.mu.Lock()
+			if _, found := api.knownDevcontainers[workspaceFolder]; !found {
+				logger.Debug(api.ctx, "adding dev container project")
+
+				dc := codersdk.WorkspaceAgentDevcontainer{
+					ID:              uuid.New(),
+					Name:            "", // Updated later based on container state.
+					WorkspaceFolder: workspaceFolder,
+					ConfigPath:      path,
+					Status:          codersdk.WorkspaceAgentDevcontainerStatusStopped,
+					Dirty:           false, // Updated later based on config file changes.
+					Container:       nil,
+				}
+
+				if api.discoveryAutostart {
+					config, err := api.dccli.ReadConfig(api.ctx, workspaceFolder, path, []string{})
+					if err != nil {
+						logger.Error(api.ctx, "read project configuration", slog.Error(err))
+					} else if config.Configuration.Customizations.Coder.AutoStart {
+						dc.Status = codersdk.WorkspaceAgentDevcontainerStatusStarting
+					}
+				}
+
+				api.knownDevcontainers[workspaceFolder] = dc
+				api.broadcastUpdatesLocked()
+
+				if dc.Status == codersdk.WorkspaceAgentDevcontainerStatusStarting {
+					api.asyncWg.Add(1)
+					go func() {
+						defer api.asyncWg.Done()
+
+						_ = api.CreateDevcontainer(dc.WorkspaceFolder, dc.ConfigPath)
+					}()
+				}
+			}
+			api.mu.Unlock()
+		}
+
+		return nil
+	})
+}
+
 func (api *API) watcherLoop() {
 	defer close(api.watcherDone)
 	defer api.logger.Debug(api.ctx, "watcher loop stopped")
@@ -449,6 +654,7 @@ func (api *API) updaterLoop() {
 	// We utilize a TickerFunc here instead of a regular Ticker so that
 	// we can guarantee execution of the updateContainers method after
 	// advancing the clock.
+	var prevErr error
 	ticker := api.clock.TickerFunc(api.ctx, api.updateInterval, func() error {
 		done := make(chan error, 1)
 		var sent bool
@@ -466,12 +672,16 @@ func (api *API) updaterLoop() {
 			if err != nil {
 				if errors.Is(err, context.Canceled) {
 					api.logger.Warn(api.ctx, "updater loop ticker canceled", slog.Error(err))
-				} else {
+					return nil
+				}
+				// Avoid excessive logging of the same error.
+				if prevErr == nil || prevErr.Error() != err.Error() {
 					api.logger.Error(api.ctx, "updater loop ticker failed", slog.Error(err))
 				}
+				prevErr = err
+			} else {
+				prevErr = nil
 			}
-		default:
-			api.logger.Debug(api.ctx, "updater loop ticker skipped, update in progress")
 		}
 
 		return nil // Always nil to keep the ticker going.
@@ -528,6 +738,7 @@ func (api *API) Routes() http.Handler {
 	r.Use(ensureInitDoneMW)
 
 	r.Get("/", api.handleList)
+	r.Get("/watch", api.watchContainers)
 	// TODO(mafredri): Simplify this route as the previous /devcontainers
 	// /-route was dropped. We can drop the /devcontainers prefix here too.
 	r.Route("/devcontainers/{devcontainer}", func(r chi.Router) {
@@ -537,6 +748,92 @@ func (api *API) Routes() http.Handler {
 	return r
 }
 
+func (api *API) broadcastUpdatesLocked() {
+	// Broadcast state changes to WebSocket listeners.
+	for _, ch := range api.updateChans {
+		select {
+		case ch <- struct{}{}:
+		default:
+		}
+	}
+}
+
+func (api *API) watchContainers(rw http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	conn, err := websocket.Accept(rw, r, &websocket.AcceptOptions{
+		// We want `NoContextTakeover` compression to balance improving
+		// bandwidth cost/latency with minimal memory usage overhead.
+		CompressionMode: websocket.CompressionNoContextTakeover,
+	})
+	if err != nil {
+		httpapi.Write(ctx, rw, http.StatusInternalServerError, codersdk.Response{
+			Message: "Failed to upgrade connection to websocket.",
+			Detail:  err.Error(),
+		})
+		return
+	}
+
+	// Here we close the websocket for reading, so that the websocket library will handle pings and
+	// close frames.
+	_ = conn.CloseRead(context.Background())
+
+	ctx, wsNetConn := codersdk.WebsocketNetConn(ctx, conn, websocket.MessageText)
+	defer wsNetConn.Close()
+
+	go httpapi.Heartbeat(ctx, conn)
+
+	updateCh := make(chan struct{}, 1)
+
+	api.mu.Lock()
+	api.updateChans = append(api.updateChans, updateCh)
+	api.mu.Unlock()
+
+	defer func() {
+		api.mu.Lock()
+		api.updateChans = slices.DeleteFunc(api.updateChans, func(ch chan struct{}) bool {
+			return ch == updateCh
+		})
+		close(updateCh)
+		api.mu.Unlock()
+	}()
+
+	encoder := json.NewEncoder(wsNetConn)
+
+	ct, err := api.getContainers()
+	if err != nil {
+		api.logger.Error(ctx, "unable to get containers", slog.Error(err))
+		return
+	}
+
+	if err := encoder.Encode(ct); err != nil {
+		api.logger.Error(ctx, "encode container list", slog.Error(err))
+		return
+	}
+
+	for {
+		select {
+		case <-api.ctx.Done():
+			return
+
+		case <-ctx.Done():
+			return
+
+		case <-updateCh:
+			ct, err := api.getContainers()
+			if err != nil {
+				api.logger.Error(ctx, "unable to get containers", slog.Error(err))
+				continue
+			}
+
+			if err := encoder.Encode(ct); err != nil {
+				api.logger.Error(ctx, "encode container list", slog.Error(err))
+				return
+			}
+		}
+	}
+}
+
 // handleList handles the HTTP request to list containers.
 func (api *API) handleList(rw http.ResponseWriter, r *http.Request) {
 	ct, err := api.getContainers()
@@ -576,8 +873,26 @@ func (api *API) updateContainers(ctx context.Context) error {
 	api.mu.Lock()
 	defer api.mu.Unlock()
 
+	var previouslyKnownDevcontainers map[string]codersdk.WorkspaceAgentDevcontainer
+	if len(api.updateChans) > 0 {
+		previouslyKnownDevcontainers = maps.Clone(api.knownDevcontainers)
+	}
+
 	api.processUpdatedContainersLocked(ctx, updated)
 
+	if len(api.updateChans) > 0 {
+		statesAreEqual := maps.EqualFunc(
+			previouslyKnownDevcontainers,
+			api.knownDevcontainers,
+			func(dc1, dc2 codersdk.WorkspaceAgentDevcontainer) bool {
+				return dc1.Equals(dc2)
+			})
+
+		if !statesAreEqual {
+			api.broadcastUpdatesLocked()
+		}
+	}
+
 	api.logger.Debug(ctx, "containers updated successfully", slog.F("container_count", len(api.containers.Containers)), slog.F("warning_count", len(api.containers.Warnings)), slog.F("devcontainer_count", len(api.knownDevcontainers)))
 
 	return nil
@@ -626,17 +941,22 @@ func (api *API) processUpdatedContainersLocked(ctx context.Context, updated code
 			slog.F("config_file", configFile),
 		)
 
+		// If we haven't set any include filters, we should explicitly ignore test devcontainers.
+		if len(api.containerLabelIncludeFilter) == 0 && container.Labels[DevcontainerIsTestRunLabel] == "true" {
+			continue
+		}
+
 		// Filter out devcontainer tests, unless explicitly set in include filters.
-		if len(api.containerLabelIncludeFilter) > 0 || container.Labels[DevcontainerIsTestRunLabel] == "true" {
-			var ok bool
+		if len(api.containerLabelIncludeFilter) > 0 {
+			includeContainer := true
 			for label, value := range api.containerLabelIncludeFilter {
-				if v, found := container.Labels[label]; found && v == value {
-					ok = true
-				}
+				v, found := container.Labels[label]
+
+				includeContainer = includeContainer && (found && v == value)
 			}
 			// Verbose debug logging is fine here since typically filters
 			// are only used in development or testing environments.
-			if !ok {
+			if !includeContainer {
 				logger.Debug(ctx, "container does not match include filter, ignoring devcontainer", slog.F("container_labels", container.Labels), slog.F("include_filter", api.containerLabelIncludeFilter))
 				continue
 			}
@@ -717,6 +1037,13 @@ func (api *API) processUpdatedContainersLocked(ctx context.Context, updated code
 				err := api.maybeInjectSubAgentIntoContainerLocked(ctx, dc)
 				if err != nil {
 					logger.Error(ctx, "inject subagent into container failed", slog.Error(err))
+					dc.Error = err.Error()
+				} else {
+					// TODO(mafredri): Preserve the error from devcontainer
+					// up if it was a lifecycle script error. Currently
+					// this results in a brief flicker for the user if
+					// injection is fast, as the error is shown then erased.
+					dc.Error = ""
 				}
 			}
 
@@ -943,7 +1270,10 @@ func (api *API) handleDevcontainerRecreate(w http.ResponseWriter, r *http.Reques
 	// devcontainer multiple times in parallel.
 	dc.Status = codersdk.WorkspaceAgentDevcontainerStatusStarting
 	dc.Container = nil
+	dc.Error = ""
 	api.knownDevcontainers[dc.WorkspaceFolder] = dc
+	api.broadcastUpdatesLocked()
+
 	go func() {
 		_ = api.CreateDevcontainer(dc.WorkspaceFolder, dc.ConfigPath, WithRemoveExistingContainer())
 	}()
@@ -1021,26 +1351,41 @@ func (api *API) CreateDevcontainer(workspaceFolder, configPath string, opts ...D
 	upOptions := []DevcontainerCLIUpOptions{WithUpOutput(infoW, errW)}
 	upOptions = append(upOptions, opts...)
 
-	_, err := api.dccli.Up(ctx, dc.WorkspaceFolder, configPath, upOptions...)
-	if err != nil {
+	containerID, upErr := api.dccli.Up(ctx, dc.WorkspaceFolder, configPath, upOptions...)
+	if upErr != nil {
 		// No need to log if the API is closing (context canceled), as this
 		// is expected behavior when the API is shutting down.
-		if !errors.Is(err, context.Canceled) {
-			logger.Error(ctx, "devcontainer creation failed", slog.Error(err))
+		if !errors.Is(upErr, context.Canceled) {
+			logger.Error(ctx, "devcontainer creation failed", slog.Error(upErr))
 		}
 
-		api.mu.Lock()
-		dc = api.knownDevcontainers[dc.WorkspaceFolder]
-		dc.Status = codersdk.WorkspaceAgentDevcontainerStatusError
-		api.knownDevcontainers[dc.WorkspaceFolder] = dc
-		api.recreateErrorTimes[dc.WorkspaceFolder] = api.clock.Now("agentcontainers", "recreate", "errorTimes")
-		api.mu.Unlock()
+		// If we don't have a container ID, the error is fatal, so we
+		// should mark the devcontainer as errored and return.
+		if containerID == "" {
+			api.mu.Lock()
+			dc = api.knownDevcontainers[dc.WorkspaceFolder]
+			dc.Status = codersdk.WorkspaceAgentDevcontainerStatusError
+			dc.Error = upErr.Error()
+			api.knownDevcontainers[dc.WorkspaceFolder] = dc
+			api.recreateErrorTimes[dc.WorkspaceFolder] = api.clock.Now("agentcontainers", "recreate", "errorTimes")
+			api.broadcastUpdatesLocked()
+			api.mu.Unlock()
 
-		return xerrors.Errorf("start devcontainer: %w", err)
+			return xerrors.Errorf("start devcontainer: %w", upErr)
+		}
+
+		// If we have a container ID, it means the container was created
+		// but a lifecycle script (e.g. postCreateCommand) failed. In this
+		// case, we still want to refresh containers to pick up the new
+		// container, inject the agent, and allow the user to debug the
+		// issue. We store the error to surface it to the user.
+		logger.Warn(ctx, "devcontainer created with errors (e.g. lifecycle script failure), container is available",
+			slog.F("container_id", containerID),
+		)
+	} else {
+		logger.Info(ctx, "devcontainer created successfully")
 	}
 
-	logger.Info(ctx, "devcontainer created successfully")
-
 	api.mu.Lock()
 	dc = api.knownDevcontainers[dc.WorkspaceFolder]
 	// Update the devcontainer status to Running or Stopped based on the
@@ -1049,14 +1394,21 @@ func (api *API) CreateDevcontainer(workspaceFolder, configPath string, opts ...D
 	// to minimize the time between API consistency, we guess the status
 	// based on the container state.
 	dc.Status = codersdk.WorkspaceAgentDevcontainerStatusStopped
-	if dc.Container != nil {
-		if dc.Container.Running {
-			dc.Status = codersdk.WorkspaceAgentDevcontainerStatusRunning
-		}
+	if dc.Container != nil && dc.Container.Running {
+		dc.Status = codersdk.WorkspaceAgentDevcontainerStatusRunning
 	}
 	dc.Dirty = false
+	if upErr != nil {
+		// If there was a lifecycle script error but we have a container ID,
+		// the container is running so we should set the status to Running.
+		dc.Status = codersdk.WorkspaceAgentDevcontainerStatusRunning
+		dc.Error = upErr.Error()
+	} else {
+		dc.Error = ""
+	}
 	api.recreateSuccessTimes[dc.WorkspaceFolder] = api.clock.Now("agentcontainers", "recreate", "successTimes")
 	api.knownDevcontainers[dc.WorkspaceFolder] = dc
+	api.broadcastUpdatesLocked()
 	api.mu.Unlock()
 
 	// Ensure an immediate refresh to accurately reflect the
@@ -1523,7 +1875,9 @@ func (api *API) maybeInjectSubAgentIntoContainerLocked(ctx context.Context, dc c
 		originalName := subAgentConfig.Name
 
 		for attempt := 1; attempt <= maxAttemptsToNameAgent; attempt++ {
-			if proc.agent, err = client.Create(ctx, subAgentConfig); err == nil {
+			agent, err := client.Create(ctx, subAgentConfig)
+			if err == nil {
+				proc.agent = agent // Only reassign on success.
 				if api.usingWorkspaceFolderName[dc.WorkspaceFolder] {
 					api.devcontainerNames[dc.Name] = true
 					delete(api.usingWorkspaceFolderName, dc.WorkspaceFolder)
@@ -1531,7 +1885,6 @@ func (api *API) maybeInjectSubAgentIntoContainerLocked(ctx context.Context, dc c
 
 				break
 			}
-
 			// NOTE(DanielleMaywood):
 			// Ordinarily we'd use `errors.As` here, but it didn't appear to work. Not
 			// sure if this is because of the communication protocol? Instead I've opted
@@ -1686,6 +2039,9 @@ func (api *API) Close() error {
 	if api.updaterDone != nil {
 		<-api.updaterDone
 	}
+	if api.discoverDone != nil {
+		<-api.discoverDone
+	}
 
 	// Wait for all async tasks to complete.
 	api.asyncWg.Wait()

agent/agentcontainers/containers_dockercli_test.go

@@ -55,11 +55,11 @@ func TestIntegrationDockerCLI(t *testing.T) {
 	}, testutil.WaitShort, testutil.IntervalSlow, "Container did not start in time")
 
 	dcli := agentcontainers.NewDockerCLI(agentexec.DefaultExecer)
-	ctx := testutil.Context(t, testutil.WaitMedium) // Longer timeout for multiple subtests
 	containerName := strings.TrimPrefix(ct.Container.Name, "/")
 
 	t.Run("DetectArchitecture", func(t *testing.T) {
 		t.Parallel()
+		ctx := testutil.Context(t, testutil.WaitShort)
 
 		arch, err := dcli.DetectArchitecture(ctx, containerName)
 		require.NoError(t, err, "DetectArchitecture failed")
@@ -71,6 +71,7 @@ func TestIntegrationDockerCLI(t *testing.T) {
 
 	t.Run("Copy", func(t *testing.T) {
 		t.Parallel()
+		ctx := testutil.Context(t, testutil.WaitShort)
 
 		want := "Help, I'm trapped!"
 		tempFile := filepath.Join(t.TempDir(), "test-file.txt")
@@ -90,6 +91,7 @@ func TestIntegrationDockerCLI(t *testing.T) {
 
 	t.Run("ExecAs", func(t *testing.T) {
 		t.Parallel()
+		ctx := testutil.Context(t, testutil.WaitShort)
 
 		// Test ExecAs without specifying user (should use container's default).
 		want := "root"

agent/agentcontainers/dcspec/gen.sh

@@ -61,7 +61,7 @@ fi
 exec 3>&-
 
 # Format the generated code.
-go run mvdan.cc/gofumpt@v0.4.0 -w -l "${TMPDIR}/${DEST_FILENAME}"
+go run mvdan.cc/gofumpt@v0.8.0 -w -l "${TMPDIR}/${DEST_FILENAME}"
 
 # Add a header so that Go recognizes this as a generated file.
 if grep -q -- "\[-i extension\]" < <(sed -h 2>&1); then

agent/agentcontainers/devcontainercli.go

@@ -91,6 +91,7 @@ type CoderCustomization struct {
 	Apps        []SubAgentApp                `json:"apps,omitempty"`
 	Name        string                       `json:"name,omitempty"`
 	Ignore      bool                         `json:"ignore,omitempty"`
+	AutoStart   bool                         `json:"autoStart,omitempty"`
 }
 
 type DevcontainerWorkspace struct {
@@ -106,63 +107,63 @@ type DevcontainerCLI interface {
 
 // DevcontainerCLIUpOptions are options for the devcontainer CLI Up
 // command.
-type DevcontainerCLIUpOptions func(*devcontainerCLIUpConfig)
+type DevcontainerCLIUpOptions func(*DevcontainerCLIUpConfig)
 
-type devcontainerCLIUpConfig struct {
-	args   []string // Additional arguments for the Up command.
-	stdout io.Writer
-	stderr io.Writer
+type DevcontainerCLIUpConfig struct {
+	Args   []string // Additional arguments for the Up command.
+	Stdout io.Writer
+	Stderr io.Writer
 }
 
 // WithRemoveExistingContainer is an option to remove the existing
 // container.
 func WithRemoveExistingContainer() DevcontainerCLIUpOptions {
-	return func(o *devcontainerCLIUpConfig) {
-		o.args = append(o.args, "--remove-existing-container")
+	return func(o *DevcontainerCLIUpConfig) {
+		o.Args = append(o.Args, "--remove-existing-container")
 	}
 }
 
 // WithUpOutput sets additional stdout and stderr writers for logs
 // during Up operations.
 func WithUpOutput(stdout, stderr io.Writer) DevcontainerCLIUpOptions {
-	return func(o *devcontainerCLIUpConfig) {
-		o.stdout = stdout
-		o.stderr = stderr
+	return func(o *DevcontainerCLIUpConfig) {
+		o.Stdout = stdout
+		o.Stderr = stderr
 	}
 }
 
 // DevcontainerCLIExecOptions are options for the devcontainer CLI Exec
 // command.
-type DevcontainerCLIExecOptions func(*devcontainerCLIExecConfig)
+type DevcontainerCLIExecOptions func(*DevcontainerCLIExecConfig)
 
-type devcontainerCLIExecConfig struct {
-	args   []string // Additional arguments for the Exec command.
-	stdout io.Writer
-	stderr io.Writer
+type DevcontainerCLIExecConfig struct {
+	Args   []string // Additional arguments for the Exec command.
+	Stdout io.Writer
+	Stderr io.Writer
 }
 
 // WithExecOutput sets additional stdout and stderr writers for logs
 // during Exec operations.
 func WithExecOutput(stdout, stderr io.Writer) DevcontainerCLIExecOptions {
-	return func(o *devcontainerCLIExecConfig) {
-		o.stdout = stdout
-		o.stderr = stderr
+	return func(o *DevcontainerCLIExecConfig) {
+		o.Stdout = stdout
+		o.Stderr = stderr
 	}
 }
 
 // WithExecContainerID sets the container ID to target a specific
 // container.
 func WithExecContainerID(id string) DevcontainerCLIExecOptions {
-	return func(o *devcontainerCLIExecConfig) {
-		o.args = append(o.args, "--container-id", id)
+	return func(o *DevcontainerCLIExecConfig) {
+		o.Args = append(o.Args, "--container-id", id)
 	}
 }
 
 // WithRemoteEnv sets environment variables for the Exec command.
 func WithRemoteEnv(env ...string) DevcontainerCLIExecOptions {
-	return func(o *devcontainerCLIExecConfig) {
+	return func(o *DevcontainerCLIExecConfig) {
 		for _, e := range env {
-			o.args = append(o.args, "--remote-env", e)
+			o.Args = append(o.Args, "--remote-env", e)
 		}
 	}
 }
@@ -185,8 +186,8 @@ func WithReadConfigOutput(stdout, stderr io.Writer) DevcontainerCLIReadConfigOpt
 	}
 }
 
-func applyDevcontainerCLIUpOptions(opts []DevcontainerCLIUpOptions) devcontainerCLIUpConfig {
-	conf := devcontainerCLIUpConfig{stdout: io.Discard, stderr: io.Discard}
+func applyDevcontainerCLIUpOptions(opts []DevcontainerCLIUpOptions) DevcontainerCLIUpConfig {
+	conf := DevcontainerCLIUpConfig{Stdout: io.Discard, Stderr: io.Discard}
 	for _, opt := range opts {
 		if opt != nil {
 			opt(&conf)
@@ -195,8 +196,8 @@ func applyDevcontainerCLIUpOptions(opts []DevcontainerCLIUpOptions) devcontainer
 	return conf
 }
 
-func applyDevcontainerCLIExecOptions(opts []DevcontainerCLIExecOptions) devcontainerCLIExecConfig {
-	conf := devcontainerCLIExecConfig{stdout: io.Discard, stderr: io.Discard}
+func applyDevcontainerCLIExecOptions(opts []DevcontainerCLIExecOptions) DevcontainerCLIExecConfig {
+	conf := DevcontainerCLIExecConfig{Stdout: io.Discard, Stderr: io.Discard}
 	for _, opt := range opts {
 		if opt != nil {
 			opt(&conf)
@@ -241,7 +242,7 @@ func (d *devcontainerCLI) Up(ctx context.Context, workspaceFolder, configPath st
 	if configPath != "" {
 		args = append(args, "--config", configPath)
 	}
-	args = append(args, conf.args...)
+	args = append(args, conf.Args...)
 	cmd := d.execer.CommandContext(ctx, "devcontainer", args...)
 
 	// Capture stdout for parsing and stream logs for both default and provided writers.
@@ -251,22 +252,25 @@ func (d *devcontainerCLI) Up(ctx context.Context, workspaceFolder, configPath st
 		&devcontainerCLILogWriter{
 			ctx:    ctx,
 			logger: logger.With(slog.F("stdout", true)),
-			writer: conf.stdout,
+			writer: conf.Stdout,
 		},
 	)
 	// Stream stderr logs and provided writer if any.
 	cmd.Stderr = &devcontainerCLILogWriter{
 		ctx:    ctx,
 		logger: logger.With(slog.F("stderr", true)),
-		writer: conf.stderr,
+		writer: conf.Stderr,
 	}
 
 	if err := cmd.Run(); err != nil {
-		_, err2 := parseDevcontainerCLILastLine[devcontainerCLIResult](ctx, logger, stdoutBuf.Bytes())
+		result, err2 := parseDevcontainerCLILastLine[devcontainerCLIResult](ctx, logger, stdoutBuf.Bytes())
 		if err2 != nil {
 			err = errors.Join(err, err2)
 		}
-		return "", err
+		// Return the container ID if available, even if there was an error.
+		// This can happen if the container was created successfully but a
+		// lifecycle script (e.g. postCreateCommand) failed.
+		return result.ContainerID, err
 	}
 
 	result, err := parseDevcontainerCLILastLine[devcontainerCLIResult](ctx, logger, stdoutBuf.Bytes())
@@ -274,6 +278,13 @@ func (d *devcontainerCLI) Up(ctx context.Context, workspaceFolder, configPath st
 		return "", err
 	}
 
+	// Check if the result indicates an error (e.g. lifecycle script failure)
+	// but still has a container ID, allowing the caller to potentially
+	// continue with the container that was created.
+	if err := result.Err(); err != nil {
+		return result.ContainerID, err
+	}
+
 	return result.ContainerID, nil
 }
 
@@ -293,17 +304,17 @@ func (d *devcontainerCLI) Exec(ctx context.Context, workspaceFolder, configPath
 	if configPath != "" {
 		args = append(args, "--config", configPath)
 	}
-	args = append(args, conf.args...)
+	args = append(args, conf.Args...)
 	args = append(args, cmd)
 	args = append(args, cmdArgs...)
 	c := d.execer.CommandContext(ctx, "devcontainer", args...)
 
-	c.Stdout = io.MultiWriter(conf.stdout, &devcontainerCLILogWriter{
+	c.Stdout = io.MultiWriter(conf.Stdout, &devcontainerCLILogWriter{
 		ctx:    ctx,
 		logger: logger.With(slog.F("stdout", true)),
 		writer: io.Discard,
 	})
-	c.Stderr = io.MultiWriter(conf.stderr, &devcontainerCLILogWriter{
+	c.Stderr = io.MultiWriter(conf.Stderr, &devcontainerCLILogWriter{
 		ctx:    ctx,
 		logger: logger.With(slog.F("stderr", true)),
 		writer: io.Discard,
@@ -393,7 +404,10 @@ func parseDevcontainerCLILastLine[T any](ctx context.Context, logger slog.Logger
 type devcontainerCLIResult struct {
 	Outcome string `json:"outcome"` // "error", "success".
 
-	// The following fields are set if outcome is success.
+	// The following fields are typically set if outcome is success, but
+	// ContainerID may also be present when outcome is error if the
+	// container was created but a lifecycle script (e.g. postCreateCommand)
+	// failed.
 	ContainerID           string `json:"containerId"`
 	RemoteUser            string `json:"remoteUser"`
 	RemoteWorkspaceFolder string `json:"remoteWorkspaceFolder"`
@@ -403,18 +417,6 @@ type devcontainerCLIResult struct {
 	Description string `json:"description"`
 }
 
-func (r *devcontainerCLIResult) UnmarshalJSON(data []byte) error {
-	type wrapperResult devcontainerCLIResult
-
-	var wrappedResult wrapperResult
-	if err := json.Unmarshal(data, &wrappedResult); err != nil {
-		return err
-	}
-
-	*r = devcontainerCLIResult(wrappedResult)
-	return r.Err()
-}
-
 func (r devcontainerCLIResult) Err() error {
 	if r.Outcome == "success" {
 		return nil

agent/agentcontainers/devcontainercli_test.go

@@ -42,56 +42,63 @@ func TestDevcontainerCLI_ArgsAndParsing(t *testing.T) {
 		t.Parallel()
 
 		tests := []struct {
-			name      string
-			logFile   string
-			workspace string
-			config    string
-			opts      []agentcontainers.DevcontainerCLIUpOptions
-			wantArgs  string
-			wantError bool
+			name            string
+			logFile         string
+			workspace       string
+			config          string
+			opts            []agentcontainers.DevcontainerCLIUpOptions
+			wantArgs        string
+			wantError       bool
+			wantContainerID bool // If true, expect a container ID even when wantError is true.
 		}{
 			{
-				name:      "success",
-				logFile:   "up.log",
-				workspace: "/test/workspace",
-				wantArgs:  "up --log-format json --workspace-folder /test/workspace",
-				wantError: false,
+				name:            "success",
+				logFile:         "up.log",
+				workspace:       "/test/workspace",
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace",
+				wantError:       false,
+				wantContainerID: true,
 			},
 			{
-				name:      "success with config",
-				logFile:   "up.log",
-				workspace: "/test/workspace",
-				config:    "/test/config.json",
-				wantArgs:  "up --log-format json --workspace-folder /test/workspace --config /test/config.json",
-				wantError: false,
+				name:            "success with config",
+				logFile:         "up.log",
+				workspace:       "/test/workspace",
+				config:          "/test/config.json",
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace --config /test/config.json",
+				wantError:       false,
+				wantContainerID: true,
 			},
 			{
-				name:      "already exists",
-				logFile:   "up-already-exists.log",
-				workspace: "/test/workspace",
-				wantArgs:  "up --log-format json --workspace-folder /test/workspace",
-				wantError: false,
+				name:            "already exists",
+				logFile:         "up-already-exists.log",
+				workspace:       "/test/workspace",
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace",
+				wantError:       false,
+				wantContainerID: true,
 			},
 			{
-				name:      "docker error",
-				logFile:   "up-error-docker.log",
-				workspace: "/test/workspace",
-				wantArgs:  "up --log-format json --workspace-folder /test/workspace",
-				wantError: true,
+				name:            "docker error",
+				logFile:         "up-error-docker.log",
+				workspace:       "/test/workspace",
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace",
+				wantError:       true,
+				wantContainerID: false,
 			},
 			{
-				name:      "bad outcome",
-				logFile:   "up-error-bad-outcome.log",
-				workspace: "/test/workspace",
-				wantArgs:  "up --log-format json --workspace-folder /test/workspace",
-				wantError: true,
+				name:            "bad outcome",
+				logFile:         "up-error-bad-outcome.log",
+				workspace:       "/test/workspace",
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace",
+				wantError:       true,
+				wantContainerID: false,
 			},
 			{
-				name:      "does not exist",
-				logFile:   "up-error-does-not-exist.log",
-				workspace: "/test/workspace",
-				wantArgs:  "up --log-format json --workspace-folder /test/workspace",
-				wantError: true,
+				name:            "does not exist",
+				logFile:         "up-error-does-not-exist.log",
+				workspace:       "/test/workspace",
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace",
+				wantError:       true,
+				wantContainerID: false,
 			},
 			{
 				name:      "with remove existing container",
@@ -100,8 +107,21 @@ func TestDevcontainerCLI_ArgsAndParsing(t *testing.T) {
 				opts: []agentcontainers.DevcontainerCLIUpOptions{
 					agentcontainers.WithRemoveExistingContainer(),
 				},
-				wantArgs:  "up --log-format json --workspace-folder /test/workspace --remove-existing-container",
-				wantError: false,
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace --remove-existing-container",
+				wantError:       false,
+				wantContainerID: true,
+			},
+			{
+				// This test verifies that when a lifecycle script like
+				// postCreateCommand fails, the CLI returns both an error
+				// and a container ID. The caller can then proceed with
+				// agent injection into the created container.
+				name:            "lifecycle script failure with container",
+				logFile:         "up-error-lifecycle-script.log",
+				workspace:       "/test/workspace",
+				wantArgs:        "up --log-format json --workspace-folder /test/workspace",
+				wantError:       true,
+				wantContainerID: true,
 			},
 		}
 
@@ -122,10 +142,13 @@ func TestDevcontainerCLI_ArgsAndParsing(t *testing.T) {
 				containerID, err := dccli.Up(ctx, tt.workspace, tt.config, tt.opts...)
 				if tt.wantError {
 					assert.Error(t, err, "want error")
-					assert.Empty(t, containerID, "expected empty container ID")
 				} else {
 					assert.NoError(t, err, "want no error")
+				}
+				if tt.wantContainerID {
 					assert.NotEmpty(t, containerID, "expected non-empty container ID")
+				} else {
+					assert.Empty(t, containerID, "expected empty container ID")
 				}
 			})
 		}
@@ -593,7 +616,7 @@ func setupDevcontainerWorkspace(t *testing.T, workspaceFolder string) string {
 	"containerEnv": {
 		"TEST_CONTAINER": "true"
 	},
-	"runArgs": ["--label", "com.coder.test=devcontainercli"]
+	"runArgs": ["--label=com.coder.test=devcontainercli", "--label=` + agentcontainers.DevcontainerIsTestRunLabel + `=true"]
 }`
 	err = os.WriteFile(configPath, []byte(content), 0o600)
 	require.NoError(t, err, "create devcontainer.json file")

agent/agentcontainers/ignore/dir.go

@@ -0,0 +1,124 @@
+package ignore
+
+import (
+	"bytes"
+	"context"
+	"errors"
+	"io/fs"
+	"os"
+	"path/filepath"
+	"strings"
+
+	"github.com/go-git/go-git/v5/plumbing/format/config"
+	"github.com/go-git/go-git/v5/plumbing/format/gitignore"
+	"github.com/spf13/afero"
+	"golang.org/x/xerrors"
+
+	"cdr.dev/slog"
+)
+
+const (
+	gitconfigFile      = ".gitconfig"
+	gitignoreFile      = ".gitignore"
+	gitInfoExcludeFile = ".git/info/exclude"
+)
+
+func FilePathToParts(path string) []string {
+	components := []string{}
+
+	if path == "" {
+		return components
+	}
+
+	for segment := range strings.SplitSeq(filepath.Clean(path), string(filepath.Separator)) {
+		if segment != "" {
+			components = append(components, segment)
+		}
+	}
+
+	return components
+}
+
+func readIgnoreFile(fileSystem afero.Fs, path, ignore string) ([]gitignore.Pattern, error) {
+	var ps []gitignore.Pattern
+
+	data, err := afero.ReadFile(fileSystem, filepath.Join(path, ignore))
+	if err != nil && !errors.Is(err, os.ErrNotExist) {
+		return nil, err
+	}
+
+	for s := range strings.SplitSeq(string(data), "\n") {
+		if !strings.HasPrefix(s, "#") && len(strings.TrimSpace(s)) > 0 {
+			ps = append(ps, gitignore.ParsePattern(s, FilePathToParts(path)))
+		}
+	}
+
+	return ps, nil
+}
+
+func ReadPatterns(ctx context.Context, logger slog.Logger, fileSystem afero.Fs, path string) ([]gitignore.Pattern, error) {
+	var ps []gitignore.Pattern
+
+	subPs, err := readIgnoreFile(fileSystem, path, gitInfoExcludeFile)
+	if err != nil {
+		return nil, err
+	}
+
+	ps = append(ps, subPs...)
+
+	if err := afero.Walk(fileSystem, path, func(path string, info fs.FileInfo, err error) error {
+		if err != nil {
+			logger.Error(ctx, "encountered error while walking for git ignore files",
+				slog.F("path", path),
+				slog.Error(err))
+			return nil
+		}
+
+		if !info.IsDir() {
+			return nil
+		}
+
+		subPs, err := readIgnoreFile(fileSystem, path, gitignoreFile)
+		if err != nil {
+			return err
+		}
+
+		ps = append(ps, subPs...)
+
+		return nil
+	}); err != nil {
+		return nil, err
+	}
+
+	return ps, nil
+}
+
+func loadPatterns(fileSystem afero.Fs, path string) ([]gitignore.Pattern, error) {
+	data, err := afero.ReadFile(fileSystem, path)
+	if err != nil && !errors.Is(err, os.ErrNotExist) {
+		return nil, err
+	}
+
+	decoder := config.NewDecoder(bytes.NewBuffer(data))
+
+	conf := config.New()
+	if err := decoder.Decode(conf); err != nil {
+		return nil, xerrors.Errorf("decode config: %w", err)
+	}
+
+	excludes := conf.Section("core").Options.Get("excludesfile")
+	if excludes == "" {
+		return nil, nil
+	}
+
+	return readIgnoreFile(fileSystem, "", excludes)
+}
+
+func LoadGlobalPatterns(fileSystem afero.Fs) ([]gitignore.Pattern, error) {
+	home, err := os.UserHomeDir()
+	if err != nil {
+		return nil, err
+	}
+
+	return loadPatterns(fileSystem, filepath.Join(home, gitconfigFile))
+}

agent/agentcontainers/ignore/dir_test.go

@@ -0,0 +1,38 @@
+package ignore_test
+
+import (
+	"fmt"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+
+	"github.com/coder/coder/v2/agent/agentcontainers/ignore"
+)
+
+func TestFilePathToParts(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		path     string
+		expected []string
+	}{
+		{"", []string{}},
+		{"/", []string{}},
+		{"foo", []string{"foo"}},
+		{"/foo", []string{"foo"}},
+		{"./foo/bar", []string{"foo", "bar"}},
+		{"../foo/bar", []string{"..", "foo", "bar"}},
+		{"foo/bar/baz", []string{"foo", "bar", "baz"}},
+		{"/foo/bar/baz", []string{"foo", "bar", "baz"}},
+		{"foo/../bar", []string{"bar"}},
+	}
+
+	for _, tt := range tests {
+		t.Run(fmt.Sprintf("`%s`", tt.path), func(t *testing.T) {
+			t.Parallel()
+
+			parts := ignore.FilePathToParts(tt.path)
+			require.Equal(t, tt.expected, parts)
+		})
+	}
+}

agent/agentcontainers/subagent.go

@@ -188,7 +188,7 @@ func (a *subAgentAPIClient) List(ctx context.Context) ([]SubAgent, error) {
 	return agents, nil
 }
 
-func (a *subAgentAPIClient) Create(ctx context.Context, agent SubAgent) (SubAgent, error) {
+func (a *subAgentAPIClient) Create(ctx context.Context, agent SubAgent) (_ SubAgent, err error) {
 	a.logger.Debug(ctx, "creating sub agent", slog.F("name", agent.Name), slog.F("directory", agent.Directory))
 
 	displayApps := make([]agentproto.CreateSubAgentRequest_DisplayApp, 0, len(agent.DisplayApps))
@@ -233,19 +233,27 @@ func (a *subAgentAPIClient) Create(ctx context.Context, agent SubAgent) (SubAgen
 	if err != nil {
 		return SubAgent{}, err
 	}
+	defer func() {
+		if err != nil {
+			// Best effort.
+			_, _ = a.api.DeleteSubAgent(ctx, &agentproto.DeleteSubAgentRequest{
+				Id: resp.GetAgent().GetId(),
+			})
+		}
+	}()
 
-	agent.Name = resp.Agent.Name
-	agent.ID, err = uuid.FromBytes(resp.Agent.Id)
+	agent.Name = resp.GetAgent().GetName()
+	agent.ID, err = uuid.FromBytes(resp.GetAgent().GetId())
 	if err != nil {
-		return agent, err
+		return SubAgent{}, err
 	}
-	agent.AuthToken, err = uuid.FromBytes(resp.Agent.AuthToken)
+	agent.AuthToken, err = uuid.FromBytes(resp.GetAgent().GetAuthToken())
 	if err != nil {
-		return agent, err
+		return SubAgent{}, err
 	}
 
-	for _, appError := range resp.AppCreationErrors {
-		app := apps[appError.Index]
+	for _, appError := range resp.GetAppCreationErrors() {
+		app := apps[appError.GetIndex()]
 
 		a.logger.Warn(ctx, "unable to create app",
 			slog.F("agent_name", agent.Name),

agent/agentcontainers/watcher/watcher_test.go

@@ -4,6 +4,7 @@ import (
 	"context"
 	"os"
 	"path/filepath"
+	"runtime"
 	"testing"
 
 	"github.com/fsnotify/fsnotify"
@@ -88,24 +89,34 @@ func TestFSNotifyWatcher(t *testing.T) {
 		break
 	}
 
-	err = os.WriteFile(testFile+".atomic", []byte(`{"test": "atomic"}`), 0o600)
-	require.NoError(t, err, "write new atomic test file failed")
+	// TODO(DanielleMaywood):
+	// Unfortunately it appears this atomic-rename phase of the test is flakey on macOS.
+	//
+	// This test flake could be indicative of an issue that may present itself
+	// in a running environment. Fortunately, we only use this (as of 2025-07-29)
+	// for our dev container integration. We do not expect the host workspace
+	// (where this is used), to ever be run on macOS, as containers are a linux
+	// paradigm.
+	if runtime.GOOS != "darwin" {
+		err = os.WriteFile(testFile+".atomic", []byte(`{"test": "atomic"}`), 0o600)
+		require.NoError(t, err, "write new atomic test file failed")
 
-	err = os.Rename(testFile+".atomic", testFile)
-	require.NoError(t, err, "rename atomic test file failed")
+		err = os.Rename(testFile+".atomic", testFile)
+		require.NoError(t, err, "rename atomic test file failed")
 
-	// Verify that we receive the event we want.
-	for {
-		event, err := wut.Next(ctx)
-		require.NoError(t, err, "next event failed")
-		require.NotNil(t, event, "want non-nil event")
-		if !event.Has(fsnotify.Create) {
-			t.Logf("Ignoring event: %s", event)
-			continue
+		// Verify that we receive the event we want.
+		for {
+			event, err := wut.Next(ctx)
+			require.NoError(t, err, "next event failed")
+			require.NotNil(t, event, "want non-nil event")
+			if !event.Has(fsnotify.Create) {
+				t.Logf("Ignoring event: %s", event)
+				continue
+			}
+			require.Truef(t, event.Has(fsnotify.Create), "want create event: %s", event.String())
+			require.Equal(t, event.Name, testFile, "want event for test file")
+			break
 		}
-		require.Truef(t, event.Has(fsnotify.Create), "want create event: %s", event.String())
-		require.Equal(t, event.Name, testFile, "want event for test file")
-		break
 	}
 
 	// Test removing the file from the watcher.

agent/agentscripts/agentscripts.go

@@ -149,7 +149,6 @@ func (r *Runner) Init(scripts []codersdk.WorkspaceAgentScript, scriptCompleted S
 		if script.Cron == "" {
 			continue
 		}
-		script := script
 		_, err := r.cron.AddFunc(script.Cron, func() {
 			err := r.trackRun(r.cronCtx, script, ExecuteCronScripts)
 			if err != nil {
@@ -224,7 +223,6 @@ func (r *Runner) Execute(ctx context.Context, option ExecuteOption) error {
 			continue
 		}
 
-		script := script
 		eg.Go(func() error {
 			err := r.trackRun(ctx, script, option)
 			if err != nil {

agent/agentsocket/client.go

@@ -0,0 +1,146 @@
+package agentsocket
+
+import (
+	"context"
+
+	"golang.org/x/xerrors"
+	"storj.io/drpc"
+	"storj.io/drpc/drpcconn"
+
+	"github.com/coder/coder/v2/agent/agentsocket/proto"
+	"github.com/coder/coder/v2/agent/unit"
+)
+
+// Option represents a configuration option for NewClient.
+type Option func(*options)
+
+type options struct {
+	path string
+}
+
+// WithPath sets the socket path. If not provided or empty, the client will
+// auto-discover the default socket path.
+func WithPath(path string) Option {
+	return func(opts *options) {
+		if path == "" {
+			return
+		}
+		opts.path = path
+	}
+}
+
+// Client provides a client for communicating with the workspace agentsocket API.
+type Client struct {
+	client proto.DRPCAgentSocketClient
+	conn   drpc.Conn
+}
+
+// NewClient creates a new socket client and opens a connection to the socket.
+// If path is not provided via WithPath or is empty, it will auto-discover the
+// default socket path.
+func NewClient(ctx context.Context, opts ...Option) (*Client, error) {
+	options := &options{}
+	for _, opt := range opts {
+		opt(options)
+	}
+
+	conn, err := dialSocket(ctx, options.path)
+	if err != nil {
+		return nil, xerrors.Errorf("connect to socket: %w", err)
+	}
+
+	drpcConn := drpcconn.New(conn)
+	client := proto.NewDRPCAgentSocketClient(drpcConn)
+
+	return &Client{
+		client: client,
+		conn:   drpcConn,
+	}, nil
+}
+
+// Close closes the socket connection.
+func (c *Client) Close() error {
+	return c.conn.Close()
+}
+
+// Ping sends a ping request to the agent.
+func (c *Client) Ping(ctx context.Context) error {
+	_, err := c.client.Ping(ctx, &proto.PingRequest{})
+	return err
+}
+
+// SyncStart starts a unit in the dependency graph.
+func (c *Client) SyncStart(ctx context.Context, unitName unit.ID) error {
+	_, err := c.client.SyncStart(ctx, &proto.SyncStartRequest{
+		Unit: string(unitName),
+	})
+	return err
+}
+
+// SyncWant declares a dependency between units.
+func (c *Client) SyncWant(ctx context.Context, unitName, dependsOn unit.ID) error {
+	_, err := c.client.SyncWant(ctx, &proto.SyncWantRequest{
+		Unit:      string(unitName),
+		DependsOn: string(dependsOn),
+	})
+	return err
+}
+
+// SyncComplete marks a unit as complete in the dependency graph.
+func (c *Client) SyncComplete(ctx context.Context, unitName unit.ID) error {
+	_, err := c.client.SyncComplete(ctx, &proto.SyncCompleteRequest{
+		Unit: string(unitName),
+	})
+	return err
+}
+
+// SyncReady requests whether a unit is ready to be started. That is, all dependencies are satisfied.
+func (c *Client) SyncReady(ctx context.Context, unitName unit.ID) (bool, error) {
+	resp, err := c.client.SyncReady(ctx, &proto.SyncReadyRequest{
+		Unit: string(unitName),
+	})
+	return resp.Ready, err
+}
+
+// SyncStatus gets the status of a unit and its dependencies.
+func (c *Client) SyncStatus(ctx context.Context, unitName unit.ID) (SyncStatusResponse, error) {
+	resp, err := c.client.SyncStatus(ctx, &proto.SyncStatusRequest{
+		Unit: string(unitName),
+	})
+	if err != nil {
+		return SyncStatusResponse{}, err
+	}
+
+	var dependencies []DependencyInfo
+	for _, dep := range resp.Dependencies {
+		dependencies = append(dependencies, DependencyInfo{
+			DependsOn:      unit.ID(dep.DependsOn),
+			RequiredStatus: unit.Status(dep.RequiredStatus),
+			CurrentStatus:  unit.Status(dep.CurrentStatus),
+			IsSatisfied:    dep.IsSatisfied,
+		})
+	}
+
+	return SyncStatusResponse{
+		UnitName:     unitName,
+		Status:       unit.Status(resp.Status),
+		IsReady:      resp.IsReady,
+		Dependencies: dependencies,
+	}, nil
+}
+
+// SyncStatusResponse contains the status information for a unit.
+type SyncStatusResponse struct {
+	UnitName     unit.ID          `table:"unit,default_sort" json:"unit_name"`
+	Status       unit.Status      `table:"status" json:"status"`
+	IsReady      bool             `table:"ready" json:"is_ready"`
+	Dependencies []DependencyInfo `table:"dependencies" json:"dependencies"`
+}
+
+// DependencyInfo contains information about a unit dependency.
+type DependencyInfo struct {
+	DependsOn      unit.ID     `table:"depends on,default_sort" json:"depends_on"`
+	RequiredStatus unit.Status `table:"required status" json:"required_status"`
+	CurrentStatus  unit.Status `table:"current status" json:"current_status"`
+	IsSatisfied    bool        `table:"satisfied" json:"is_satisfied"`
+}

agent/agentsocket/proto/agentsocket.pb.go

@@ -0,0 +1,968 @@
+// Code generated by protoc-gen-go. DO NOT EDIT.
+// versions:
+// 	protoc-gen-go v1.30.0
+// 	protoc        v4.23.4
+// source: agent/agentsocket/proto/agentsocket.proto
+
+package proto
+
+import (
+	protoreflect "google.golang.org/protobuf/reflect/protoreflect"
+	protoimpl "google.golang.org/protobuf/runtime/protoimpl"
+	reflect "reflect"
+	sync "sync"
+)
+
+const (
+	// Verify that this generated code is sufficiently up-to-date.
+	_ = protoimpl.EnforceVersion(20 - protoimpl.MinVersion)
+	// Verify that runtime/protoimpl is sufficiently up-to-date.
+	_ = protoimpl.EnforceVersion(protoimpl.MaxVersion - 20)
+)
+
+type PingRequest struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+}
+
+func (x *PingRequest) Reset() {
+	*x = PingRequest{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[0]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *PingRequest) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*PingRequest) ProtoMessage() {}
+
+func (x *PingRequest) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[0]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use PingRequest.ProtoReflect.Descriptor instead.
+func (*PingRequest) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{0}
+}
+
+type PingResponse struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+}
+
+func (x *PingResponse) Reset() {
+	*x = PingResponse{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[1]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *PingResponse) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*PingResponse) ProtoMessage() {}
+
+func (x *PingResponse) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[1]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use PingResponse.ProtoReflect.Descriptor instead.
+func (*PingResponse) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{1}
+}
+
+type SyncStartRequest struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Unit string `protobuf:"bytes,1,opt,name=unit,proto3" json:"unit,omitempty"`
+}
+
+func (x *SyncStartRequest) Reset() {
+	*x = SyncStartRequest{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[2]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncStartRequest) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncStartRequest) ProtoMessage() {}
+
+func (x *SyncStartRequest) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[2]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncStartRequest.ProtoReflect.Descriptor instead.
+func (*SyncStartRequest) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{2}
+}
+
+func (x *SyncStartRequest) GetUnit() string {
+	if x != nil {
+		return x.Unit
+	}
+	return ""
+}
+
+type SyncStartResponse struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+}
+
+func (x *SyncStartResponse) Reset() {
+	*x = SyncStartResponse{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[3]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncStartResponse) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncStartResponse) ProtoMessage() {}
+
+func (x *SyncStartResponse) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[3]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncStartResponse.ProtoReflect.Descriptor instead.
+func (*SyncStartResponse) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{3}
+}
+
+type SyncWantRequest struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Unit      string `protobuf:"bytes,1,opt,name=unit,proto3" json:"unit,omitempty"`
+	DependsOn string `protobuf:"bytes,2,opt,name=depends_on,json=dependsOn,proto3" json:"depends_on,omitempty"`
+}
+
+func (x *SyncWantRequest) Reset() {
+	*x = SyncWantRequest{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[4]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncWantRequest) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncWantRequest) ProtoMessage() {}
+
+func (x *SyncWantRequest) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[4]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncWantRequest.ProtoReflect.Descriptor instead.
+func (*SyncWantRequest) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{4}
+}
+
+func (x *SyncWantRequest) GetUnit() string {
+	if x != nil {
+		return x.Unit
+	}
+	return ""
+}
+
+func (x *SyncWantRequest) GetDependsOn() string {
+	if x != nil {
+		return x.DependsOn
+	}
+	return ""
+}
+
+type SyncWantResponse struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+}
+
+func (x *SyncWantResponse) Reset() {
+	*x = SyncWantResponse{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[5]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncWantResponse) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncWantResponse) ProtoMessage() {}
+
+func (x *SyncWantResponse) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[5]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncWantResponse.ProtoReflect.Descriptor instead.
+func (*SyncWantResponse) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{5}
+}
+
+type SyncCompleteRequest struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Unit string `protobuf:"bytes,1,opt,name=unit,proto3" json:"unit,omitempty"`
+}
+
+func (x *SyncCompleteRequest) Reset() {
+	*x = SyncCompleteRequest{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[6]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncCompleteRequest) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncCompleteRequest) ProtoMessage() {}
+
+func (x *SyncCompleteRequest) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[6]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncCompleteRequest.ProtoReflect.Descriptor instead.
+func (*SyncCompleteRequest) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{6}
+}
+
+func (x *SyncCompleteRequest) GetUnit() string {
+	if x != nil {
+		return x.Unit
+	}
+	return ""
+}
+
+type SyncCompleteResponse struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+}
+
+func (x *SyncCompleteResponse) Reset() {
+	*x = SyncCompleteResponse{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[7]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncCompleteResponse) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncCompleteResponse) ProtoMessage() {}
+
+func (x *SyncCompleteResponse) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[7]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncCompleteResponse.ProtoReflect.Descriptor instead.
+func (*SyncCompleteResponse) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{7}
+}
+
+type SyncReadyRequest struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Unit string `protobuf:"bytes,1,opt,name=unit,proto3" json:"unit,omitempty"`
+}
+
+func (x *SyncReadyRequest) Reset() {
+	*x = SyncReadyRequest{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[8]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncReadyRequest) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncReadyRequest) ProtoMessage() {}
+
+func (x *SyncReadyRequest) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[8]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncReadyRequest.ProtoReflect.Descriptor instead.
+func (*SyncReadyRequest) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{8}
+}
+
+func (x *SyncReadyRequest) GetUnit() string {
+	if x != nil {
+		return x.Unit
+	}
+	return ""
+}
+
+type SyncReadyResponse struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Ready bool `protobuf:"varint,1,opt,name=ready,proto3" json:"ready,omitempty"`
+}
+
+func (x *SyncReadyResponse) Reset() {
+	*x = SyncReadyResponse{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[9]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncReadyResponse) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncReadyResponse) ProtoMessage() {}
+
+func (x *SyncReadyResponse) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[9]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncReadyResponse.ProtoReflect.Descriptor instead.
+func (*SyncReadyResponse) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{9}
+}
+
+func (x *SyncReadyResponse) GetReady() bool {
+	if x != nil {
+		return x.Ready
+	}
+	return false
+}
+
+type SyncStatusRequest struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Unit string `protobuf:"bytes,1,opt,name=unit,proto3" json:"unit,omitempty"`
+}
+
+func (x *SyncStatusRequest) Reset() {
+	*x = SyncStatusRequest{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[10]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncStatusRequest) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncStatusRequest) ProtoMessage() {}
+
+func (x *SyncStatusRequest) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[10]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncStatusRequest.ProtoReflect.Descriptor instead.
+func (*SyncStatusRequest) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{10}
+}
+
+func (x *SyncStatusRequest) GetUnit() string {
+	if x != nil {
+		return x.Unit
+	}
+	return ""
+}
+
+type DependencyInfo struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Unit           string `protobuf:"bytes,1,opt,name=unit,proto3" json:"unit,omitempty"`
+	DependsOn      string `protobuf:"bytes,2,opt,name=depends_on,json=dependsOn,proto3" json:"depends_on,omitempty"`
+	RequiredStatus string `protobuf:"bytes,3,opt,name=required_status,json=requiredStatus,proto3" json:"required_status,omitempty"`
+	CurrentStatus  string `protobuf:"bytes,4,opt,name=current_status,json=currentStatus,proto3" json:"current_status,omitempty"`
+	IsSatisfied    bool   `protobuf:"varint,5,opt,name=is_satisfied,json=isSatisfied,proto3" json:"is_satisfied,omitempty"`
+}
+
+func (x *DependencyInfo) Reset() {
+	*x = DependencyInfo{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[11]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *DependencyInfo) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*DependencyInfo) ProtoMessage() {}
+
+func (x *DependencyInfo) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[11]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use DependencyInfo.ProtoReflect.Descriptor instead.
+func (*DependencyInfo) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{11}
+}
+
+func (x *DependencyInfo) GetUnit() string {
+	if x != nil {
+		return x.Unit
+	}
+	return ""
+}
+
+func (x *DependencyInfo) GetDependsOn() string {
+	if x != nil {
+		return x.DependsOn
+	}
+	return ""
+}
+
+func (x *DependencyInfo) GetRequiredStatus() string {
+	if x != nil {
+		return x.RequiredStatus
+	}
+	return ""
+}
+
+func (x *DependencyInfo) GetCurrentStatus() string {
+	if x != nil {
+		return x.CurrentStatus
+	}
+	return ""
+}
+
+func (x *DependencyInfo) GetIsSatisfied() bool {
+	if x != nil {
+		return x.IsSatisfied
+	}
+	return false
+}
+
+type SyncStatusResponse struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Status       string            `protobuf:"bytes,1,opt,name=status,proto3" json:"status,omitempty"`
+	IsReady      bool              `protobuf:"varint,2,opt,name=is_ready,json=isReady,proto3" json:"is_ready,omitempty"`
+	Dependencies []*DependencyInfo `protobuf:"bytes,3,rep,name=dependencies,proto3" json:"dependencies,omitempty"`
+}
+
+func (x *SyncStatusResponse) Reset() {
+	*x = SyncStatusResponse{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[12]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *SyncStatusResponse) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*SyncStatusResponse) ProtoMessage() {}
+
+func (x *SyncStatusResponse) ProtoReflect() protoreflect.Message {
+	mi := &file_agent_agentsocket_proto_agentsocket_proto_msgTypes[12]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use SyncStatusResponse.ProtoReflect.Descriptor instead.
+func (*SyncStatusResponse) Descriptor() ([]byte, []int) {
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP(), []int{12}
+}
+
+func (x *SyncStatusResponse) GetStatus() string {
+	if x != nil {
+		return x.Status
+	}
+	return ""
+}
+
+func (x *SyncStatusResponse) GetIsReady() bool {
+	if x != nil {
+		return x.IsReady
+	}
+	return false
+}
+
+func (x *SyncStatusResponse) GetDependencies() []*DependencyInfo {
+	if x != nil {
+		return x.Dependencies
+	}
+	return nil
+}
+
+var File_agent_agentsocket_proto_agentsocket_proto protoreflect.FileDescriptor
+
+var file_agent_agentsocket_proto_agentsocket_proto_rawDesc = []byte{
+	0x0a, 0x29, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x2f, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63,
+	0x6b, 0x65, 0x74, 0x2f, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x2f, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73,
+	0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x12, 0x14, 0x63, 0x6f, 0x64,
+	0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e, 0x76,
+	0x31, 0x22, 0x0d, 0x0a, 0x0b, 0x50, 0x69, 0x6e, 0x67, 0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74,
+	0x22, 0x0e, 0x0a, 0x0c, 0x50, 0x69, 0x6e, 0x67, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65,
+	0x22, 0x26, 0x0a, 0x10, 0x53, 0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x72, 0x74, 0x52, 0x65, 0x71,
+	0x75, 0x65, 0x73, 0x74, 0x12, 0x12, 0x0a, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x18, 0x01, 0x20, 0x01,
+	0x28, 0x09, 0x52, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x22, 0x13, 0x0a, 0x11, 0x53, 0x79, 0x6e, 0x63,
+	0x53, 0x74, 0x61, 0x72, 0x74, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65, 0x22, 0x44, 0x0a,
+	0x0f, 0x53, 0x79, 0x6e, 0x63, 0x57, 0x61, 0x6e, 0x74, 0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74,
+	0x12, 0x12, 0x0a, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x04,
+	0x75, 0x6e, 0x69, 0x74, 0x12, 0x1d, 0x0a, 0x0a, 0x64, 0x65, 0x70, 0x65, 0x6e, 0x64, 0x73, 0x5f,
+	0x6f, 0x6e, 0x18, 0x02, 0x20, 0x01, 0x28, 0x09, 0x52, 0x09, 0x64, 0x65, 0x70, 0x65, 0x6e, 0x64,
+	0x73, 0x4f, 0x6e, 0x22, 0x12, 0x0a, 0x10, 0x53, 0x79, 0x6e, 0x63, 0x57, 0x61, 0x6e, 0x74, 0x52,
+	0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65, 0x22, 0x29, 0x0a, 0x13, 0x53, 0x79, 0x6e, 0x63, 0x43,
+	0x6f, 0x6d, 0x70, 0x6c, 0x65, 0x74, 0x65, 0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74, 0x12, 0x12,
+	0x0a, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x04, 0x75, 0x6e,
+	0x69, 0x74, 0x22, 0x16, 0x0a, 0x14, 0x53, 0x79, 0x6e, 0x63, 0x43, 0x6f, 0x6d, 0x70, 0x6c, 0x65,
+	0x74, 0x65, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65, 0x22, 0x26, 0x0a, 0x10, 0x53, 0x79,
+	0x6e, 0x63, 0x52, 0x65, 0x61, 0x64, 0x79, 0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74, 0x12, 0x12,
+	0x0a, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x04, 0x75, 0x6e,
+	0x69, 0x74, 0x22, 0x29, 0x0a, 0x11, 0x53, 0x79, 0x6e, 0x63, 0x52, 0x65, 0x61, 0x64, 0x79, 0x52,
+	0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65, 0x12, 0x14, 0x0a, 0x05, 0x72, 0x65, 0x61, 0x64, 0x79,
+	0x18, 0x01, 0x20, 0x01, 0x28, 0x08, 0x52, 0x05, 0x72, 0x65, 0x61, 0x64, 0x79, 0x22, 0x27, 0x0a,
+	0x11, 0x53, 0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x74, 0x75, 0x73, 0x52, 0x65, 0x71, 0x75, 0x65,
+	0x73, 0x74, 0x12, 0x12, 0x0a, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09,
+	0x52, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x22, 0xb6, 0x01, 0x0a, 0x0e, 0x44, 0x65, 0x70, 0x65, 0x6e,
+	0x64, 0x65, 0x6e, 0x63, 0x79, 0x49, 0x6e, 0x66, 0x6f, 0x12, 0x12, 0x0a, 0x04, 0x75, 0x6e, 0x69,
+	0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x04, 0x75, 0x6e, 0x69, 0x74, 0x12, 0x1d, 0x0a,
+	0x0a, 0x64, 0x65, 0x70, 0x65, 0x6e, 0x64, 0x73, 0x5f, 0x6f, 0x6e, 0x18, 0x02, 0x20, 0x01, 0x28,
+	0x09, 0x52, 0x09, 0x64, 0x65, 0x70, 0x65, 0x6e, 0x64, 0x73, 0x4f, 0x6e, 0x12, 0x27, 0x0a, 0x0f,
+	0x72, 0x65, 0x71, 0x75, 0x69, 0x72, 0x65, 0x64, 0x5f, 0x73, 0x74, 0x61, 0x74, 0x75, 0x73, 0x18,
+	0x03, 0x20, 0x01, 0x28, 0x09, 0x52, 0x0e, 0x72, 0x65, 0x71, 0x75, 0x69, 0x72, 0x65, 0x64, 0x53,
+	0x74, 0x61, 0x74, 0x75, 0x73, 0x12, 0x25, 0x0a, 0x0e, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74,
+	0x5f, 0x73, 0x74, 0x61, 0x74, 0x75, 0x73, 0x18, 0x04, 0x20, 0x01, 0x28, 0x09, 0x52, 0x0d, 0x63,
+	0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x53, 0x74, 0x61, 0x74, 0x75, 0x73, 0x12, 0x21, 0x0a, 0x0c,
+	0x69, 0x73, 0x5f, 0x73, 0x61, 0x74, 0x69, 0x73, 0x66, 0x69, 0x65, 0x64, 0x18, 0x05, 0x20, 0x01,
+	0x28, 0x08, 0x52, 0x0b, 0x69, 0x73, 0x53, 0x61, 0x74, 0x69, 0x73, 0x66, 0x69, 0x65, 0x64, 0x22,
+	0x91, 0x01, 0x0a, 0x12, 0x53, 0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x74, 0x75, 0x73, 0x52, 0x65,
+	0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65, 0x12, 0x16, 0x0a, 0x06, 0x73, 0x74, 0x61, 0x74, 0x75, 0x73,
+	0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x06, 0x73, 0x74, 0x61, 0x74, 0x75, 0x73, 0x12, 0x19,
+	0x0a, 0x08, 0x69, 0x73, 0x5f, 0x72, 0x65, 0x61, 0x64, 0x79, 0x18, 0x02, 0x20, 0x01, 0x28, 0x08,
+	0x52, 0x07, 0x69, 0x73, 0x52, 0x65, 0x61, 0x64, 0x79, 0x12, 0x48, 0x0a, 0x0c, 0x64, 0x65, 0x70,
+	0x65, 0x6e, 0x64, 0x65, 0x6e, 0x63, 0x69, 0x65, 0x73, 0x18, 0x03, 0x20, 0x03, 0x28, 0x0b, 0x32,
+	0x24, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63,
+	0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x44, 0x65, 0x70, 0x65, 0x6e, 0x64, 0x65, 0x6e, 0x63,
+	0x79, 0x49, 0x6e, 0x66, 0x6f, 0x52, 0x0c, 0x64, 0x65, 0x70, 0x65, 0x6e, 0x64, 0x65, 0x6e, 0x63,
+	0x69, 0x65, 0x73, 0x32, 0xbb, 0x04, 0x0a, 0x0b, 0x41, 0x67, 0x65, 0x6e, 0x74, 0x53, 0x6f, 0x63,
+	0x6b, 0x65, 0x74, 0x12, 0x4d, 0x0a, 0x04, 0x50, 0x69, 0x6e, 0x67, 0x12, 0x21, 0x2e, 0x63, 0x6f,
+	0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e,
+	0x76, 0x31, 0x2e, 0x50, 0x69, 0x6e, 0x67, 0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74, 0x1a, 0x22,
+	0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b,
+	0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x50, 0x69, 0x6e, 0x67, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e,
+	0x73, 0x65, 0x12, 0x5c, 0x0a, 0x09, 0x53, 0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x72, 0x74, 0x12,
+	0x26, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63,
+	0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53, 0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x72, 0x74,
+	0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74, 0x1a, 0x27, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e,
+	0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53,
+	0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x72, 0x74, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65,
+	0x12, 0x59, 0x0a, 0x08, 0x53, 0x79, 0x6e, 0x63, 0x57, 0x61, 0x6e, 0x74, 0x12, 0x25, 0x2e, 0x63,
+	0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74,
+	0x2e, 0x76, 0x31, 0x2e, 0x53, 0x79, 0x6e, 0x63, 0x57, 0x61, 0x6e, 0x74, 0x52, 0x65, 0x71, 0x75,
+	0x65, 0x73, 0x74, 0x1a, 0x26, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e,
+	0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53, 0x79, 0x6e, 0x63, 0x57,
+	0x61, 0x6e, 0x74, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65, 0x12, 0x65, 0x0a, 0x0c, 0x53,
+	0x79, 0x6e, 0x63, 0x43, 0x6f, 0x6d, 0x70, 0x6c, 0x65, 0x74, 0x65, 0x12, 0x29, 0x2e, 0x63, 0x6f,
+	0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e,
+	0x76, 0x31, 0x2e, 0x53, 0x79, 0x6e, 0x63, 0x43, 0x6f, 0x6d, 0x70, 0x6c, 0x65, 0x74, 0x65, 0x52,
+	0x65, 0x71, 0x75, 0x65, 0x73, 0x74, 0x1a, 0x2a, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61,
+	0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53, 0x79,
+	0x6e, 0x63, 0x43, 0x6f, 0x6d, 0x70, 0x6c, 0x65, 0x74, 0x65, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e,
+	0x73, 0x65, 0x12, 0x5c, 0x0a, 0x09, 0x53, 0x79, 0x6e, 0x63, 0x52, 0x65, 0x61, 0x64, 0x79, 0x12,
+	0x26, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63,
+	0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53, 0x79, 0x6e, 0x63, 0x52, 0x65, 0x61, 0x64, 0x79,
+	0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74, 0x1a, 0x27, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e,
+	0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53,
+	0x79, 0x6e, 0x63, 0x52, 0x65, 0x61, 0x64, 0x79, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73, 0x65,
+	0x12, 0x5f, 0x0a, 0x0a, 0x53, 0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x74, 0x75, 0x73, 0x12, 0x27,
+	0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b,
+	0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53, 0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x74, 0x75, 0x73,
+	0x52, 0x65, 0x71, 0x75, 0x65, 0x73, 0x74, 0x1a, 0x28, 0x2e, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2e,
+	0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74, 0x2e, 0x76, 0x31, 0x2e, 0x53,
+	0x79, 0x6e, 0x63, 0x53, 0x74, 0x61, 0x74, 0x75, 0x73, 0x52, 0x65, 0x73, 0x70, 0x6f, 0x6e, 0x73,
+	0x65, 0x42, 0x33, 0x5a, 0x31, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f,
+	0x63, 0x6f, 0x64, 0x65, 0x72, 0x2f, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x2f, 0x76, 0x32, 0x2f, 0x61,
+	0x67, 0x65, 0x6e, 0x74, 0x2f, 0x61, 0x67, 0x65, 0x6e, 0x74, 0x73, 0x6f, 0x63, 0x6b, 0x65, 0x74,
+	0x2f, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
+}
+
+var (
+	file_agent_agentsocket_proto_agentsocket_proto_rawDescOnce sync.Once
+	file_agent_agentsocket_proto_agentsocket_proto_rawDescData = file_agent_agentsocket_proto_agentsocket_proto_rawDesc
+)
+
+func file_agent_agentsocket_proto_agentsocket_proto_rawDescGZIP() []byte {
+	file_agent_agentsocket_proto_agentsocket_proto_rawDescOnce.Do(func() {
+		file_agent_agentsocket_proto_agentsocket_proto_rawDescData = protoimpl.X.CompressGZIP(file_agent_agentsocket_proto_agentsocket_proto_rawDescData)
+	})
+	return file_agent_agentsocket_proto_agentsocket_proto_rawDescData
+}
+
+var file_agent_agentsocket_proto_agentsocket_proto_msgTypes = make([]protoimpl.MessageInfo, 13)
+var file_agent_agentsocket_proto_agentsocket_proto_goTypes = []interface{}{
+	(*PingRequest)(nil),          // 0: coder.agentsocket.v1.PingRequest
+	(*PingResponse)(nil),         // 1: coder.agentsocket.v1.PingResponse
+	(*SyncStartRequest)(nil),     // 2: coder.agentsocket.v1.SyncStartRequest
+	(*SyncStartResponse)(nil),    // 3: coder.agentsocket.v1.SyncStartResponse
+	(*SyncWantRequest)(nil),      // 4: coder.agentsocket.v1.SyncWantRequest
+	(*SyncWantResponse)(nil),     // 5: coder.agentsocket.v1.SyncWantResponse
+	(*SyncCompleteRequest)(nil),  // 6: coder.agentsocket.v1.SyncCompleteRequest
+	(*SyncCompleteResponse)(nil), // 7: coder.agentsocket.v1.SyncCompleteResponse
+	(*SyncReadyRequest)(nil),     // 8: coder.agentsocket.v1.SyncReadyRequest
+	(*SyncReadyResponse)(nil),    // 9: coder.agentsocket.v1.SyncReadyResponse
+	(*SyncStatusRequest)(nil),    // 10: coder.agentsocket.v1.SyncStatusRequest
+	(*DependencyInfo)(nil),       // 11: coder.agentsocket.v1.DependencyInfo
+	(*SyncStatusResponse)(nil),   // 12: coder.agentsocket.v1.SyncStatusResponse
+}
+var file_agent_agentsocket_proto_agentsocket_proto_depIdxs = []int32{
+	11, // 0: coder.agentsocket.v1.SyncStatusResponse.dependencies:type_name -> coder.agentsocket.v1.DependencyInfo
+	0,  // 1: coder.agentsocket.v1.AgentSocket.Ping:input_type -> coder.agentsocket.v1.PingRequest
+	2,  // 2: coder.agentsocket.v1.AgentSocket.SyncStart:input_type -> coder.agentsocket.v1.SyncStartRequest
+	4,  // 3: coder.agentsocket.v1.AgentSocket.SyncWant:input_type -> coder.agentsocket.v1.SyncWantRequest
+	6,  // 4: coder.agentsocket.v1.AgentSocket.SyncComplete:input_type -> coder.agentsocket.v1.SyncCompleteRequest
+	8,  // 5: coder.agentsocket.v1.AgentSocket.SyncReady:input_type -> coder.agentsocket.v1.SyncReadyRequest
+	10, // 6: coder.agentsocket.v1.AgentSocket.SyncStatus:input_type -> coder.agentsocket.v1.SyncStatusRequest
+	1,  // 7: coder.agentsocket.v1.AgentSocket.Ping:output_type -> coder.agentsocket.v1.PingResponse
+	3,  // 8: coder.agentsocket.v1.AgentSocket.SyncStart:output_type -> coder.agentsocket.v1.SyncStartResponse
+	5,  // 9: coder.agentsocket.v1.AgentSocket.SyncWant:output_type -> coder.agentsocket.v1.SyncWantResponse
+	7,  // 10: coder.agentsocket.v1.AgentSocket.SyncComplete:output_type -> coder.agentsocket.v1.SyncCompleteResponse
+	9,  // 11: coder.agentsocket.v1.AgentSocket.SyncReady:output_type -> coder.agentsocket.v1.SyncReadyResponse
+	12, // 12: coder.agentsocket.v1.AgentSocket.SyncStatus:output_type -> coder.agentsocket.v1.SyncStatusResponse
+	7,  // [7:13] is the sub-list for method output_type
+	1,  // [1:7] is the sub-list for method input_type
+	1,  // [1:1] is the sub-list for extension type_name
+	1,  // [1:1] is the sub-list for extension extendee
+	0,  // [0:1] is the sub-list for field type_name
+}
+
+func init() { file_agent_agentsocket_proto_agentsocket_proto_init() }
+func file_agent_agentsocket_proto_agentsocket_proto_init() {
+	if File_agent_agentsocket_proto_agentsocket_proto != nil {
+		return
+	}
+	if !protoimpl.UnsafeEnabled {
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[0].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*PingRequest); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[1].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*PingResponse); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[2].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncStartRequest); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[3].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncStartResponse); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[4].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncWantRequest); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[5].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncWantResponse); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[6].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncCompleteRequest); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[7].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncCompleteResponse); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[8].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncReadyRequest); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[9].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncReadyResponse); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[10].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncStatusRequest); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[11].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*DependencyInfo); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_agent_agentsocket_proto_agentsocket_proto_msgTypes[12].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*SyncStatusResponse); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+	}
+	type x struct{}
+	out := protoimpl.TypeBuilder{
+		File: protoimpl.DescBuilder{
+			GoPackagePath: reflect.TypeOf(x{}).PkgPath(),
+			RawDescriptor: file_agent_agentsocket_proto_agentsocket_proto_rawDesc,
+			NumEnums:      0,
+			NumMessages:   13,
+			NumExtensions: 0,
+			NumServices:   1,
+		},
+		GoTypes:           file_agent_agentsocket_proto_agentsocket_proto_goTypes,
+		DependencyIndexes: file_agent_agentsocket_proto_agentsocket_proto_depIdxs,
+		MessageInfos:      file_agent_agentsocket_proto_agentsocket_proto_msgTypes,
+	}.Build()
+	File_agent_agentsocket_proto_agentsocket_proto = out.File
+	file_agent_agentsocket_proto_agentsocket_proto_rawDesc = nil
+	file_agent_agentsocket_proto_agentsocket_proto_goTypes = nil
+	file_agent_agentsocket_proto_agentsocket_proto_depIdxs = nil
+}

agent/agentsocket/proto/agentsocket.proto

@@ -0,0 +1,69 @@
+syntax = "proto3";
+option go_package = "github.com/coder/coder/v2/agent/agentsocket/proto";
+
+package coder.agentsocket.v1;
+
+message PingRequest {}
+
+message PingResponse {}
+
+message SyncStartRequest {
+  string unit = 1;
+}
+
+message SyncStartResponse {}
+
+message SyncWantRequest {
+  string unit = 1;
+  string depends_on = 2;
+}
+
+message SyncWantResponse {}
+
+message SyncCompleteRequest {
+  string unit = 1;
+}
+
+message SyncCompleteResponse {}
+
+message SyncReadyRequest {
+  string unit = 1;
+}
+
+message SyncReadyResponse {
+  bool ready = 1;
+}
+
+message SyncStatusRequest {
+  string unit = 1;
+}
+
+message DependencyInfo {
+  string unit = 1;
+  string depends_on = 2;
+  string required_status = 3;
+  string current_status = 4;
+  bool is_satisfied = 5;
+}
+
+message SyncStatusResponse {
+  string status = 1;
+  bool is_ready = 2;
+  repeated DependencyInfo dependencies = 3;
+}
+
+// AgentSocket provides direct access to the agent over local IPC.
+service AgentSocket {
+  // Ping the agent to check if it is alive.
+  rpc Ping(PingRequest) returns (PingResponse);
+  // Report the start of a unit.
+  rpc SyncStart(SyncStartRequest) returns (SyncStartResponse);
+  // Declare a dependency between units.
+  rpc SyncWant(SyncWantRequest) returns (SyncWantResponse);
+  // Report the completion of a unit.
+  rpc SyncComplete(SyncCompleteRequest) returns (SyncCompleteResponse);
+  // Request whether a unit is ready to be started. That is, all dependencies are satisfied.
+  rpc SyncReady(SyncReadyRequest) returns (SyncReadyResponse);
+  // Get the status of a unit and list its dependencies.
+  rpc SyncStatus(SyncStatusRequest) returns (SyncStatusResponse);
+}

agent/agentsocket/proto/agentsocket_drpc.pb.go

@@ -0,0 +1,311 @@
+// Code generated by protoc-gen-go-drpc. DO NOT EDIT.
+// protoc-gen-go-drpc version: v0.0.34
+// source: agent/agentsocket/proto/agentsocket.proto
+
+package proto
+
+import (
+	context "context"
+	errors "errors"
+	protojson "google.golang.org/protobuf/encoding/protojson"
+	proto "google.golang.org/protobuf/proto"
+	drpc "storj.io/drpc"
+	drpcerr "storj.io/drpc/drpcerr"
+)
+
+type drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto struct{}
+
+func (drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto) Marshal(msg drpc.Message) ([]byte, error) {
+	return proto.Marshal(msg.(proto.Message))
+}
+
+func (drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto) MarshalAppend(buf []byte, msg drpc.Message) ([]byte, error) {
+	return proto.MarshalOptions{}.MarshalAppend(buf, msg.(proto.Message))
+}
+
+func (drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto) Unmarshal(buf []byte, msg drpc.Message) error {
+	return proto.Unmarshal(buf, msg.(proto.Message))
+}
+
+func (drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto) JSONMarshal(msg drpc.Message) ([]byte, error) {
+	return protojson.Marshal(msg.(proto.Message))
+}
+
+func (drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto) JSONUnmarshal(buf []byte, msg drpc.Message) error {
+	return protojson.Unmarshal(buf, msg.(proto.Message))
+}
+
+type DRPCAgentSocketClient interface {
+	DRPCConn() drpc.Conn
+
+	Ping(ctx context.Context, in *PingRequest) (*PingResponse, error)
+	SyncStart(ctx context.Context, in *SyncStartRequest) (*SyncStartResponse, error)
+	SyncWant(ctx context.Context, in *SyncWantRequest) (*SyncWantResponse, error)
+	SyncComplete(ctx context.Context, in *SyncCompleteRequest) (*SyncCompleteResponse, error)
+	SyncReady(ctx context.Context, in *SyncReadyRequest) (*SyncReadyResponse, error)
+	SyncStatus(ctx context.Context, in *SyncStatusRequest) (*SyncStatusResponse, error)
+}
+
+type drpcAgentSocketClient struct {
+	cc drpc.Conn
+}
+
+func NewDRPCAgentSocketClient(cc drpc.Conn) DRPCAgentSocketClient {
+	return &drpcAgentSocketClient{cc}
+}
+
+func (c *drpcAgentSocketClient) DRPCConn() drpc.Conn { return c.cc }
+
+func (c *drpcAgentSocketClient) Ping(ctx context.Context, in *PingRequest) (*PingResponse, error) {
+	out := new(PingResponse)
+	err := c.cc.Invoke(ctx, "/coder.agentsocket.v1.AgentSocket/Ping", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}, in, out)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *drpcAgentSocketClient) SyncStart(ctx context.Context, in *SyncStartRequest) (*SyncStartResponse, error) {
+	out := new(SyncStartResponse)
+	err := c.cc.Invoke(ctx, "/coder.agentsocket.v1.AgentSocket/SyncStart", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}, in, out)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *drpcAgentSocketClient) SyncWant(ctx context.Context, in *SyncWantRequest) (*SyncWantResponse, error) {
+	out := new(SyncWantResponse)
+	err := c.cc.Invoke(ctx, "/coder.agentsocket.v1.AgentSocket/SyncWant", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}, in, out)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *drpcAgentSocketClient) SyncComplete(ctx context.Context, in *SyncCompleteRequest) (*SyncCompleteResponse, error) {
+	out := new(SyncCompleteResponse)
+	err := c.cc.Invoke(ctx, "/coder.agentsocket.v1.AgentSocket/SyncComplete", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}, in, out)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *drpcAgentSocketClient) SyncReady(ctx context.Context, in *SyncReadyRequest) (*SyncReadyResponse, error) {
+	out := new(SyncReadyResponse)
+	err := c.cc.Invoke(ctx, "/coder.agentsocket.v1.AgentSocket/SyncReady", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}, in, out)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *drpcAgentSocketClient) SyncStatus(ctx context.Context, in *SyncStatusRequest) (*SyncStatusResponse, error) {
+	out := new(SyncStatusResponse)
+	err := c.cc.Invoke(ctx, "/coder.agentsocket.v1.AgentSocket/SyncStatus", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}, in, out)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+type DRPCAgentSocketServer interface {
+	Ping(context.Context, *PingRequest) (*PingResponse, error)
+	SyncStart(context.Context, *SyncStartRequest) (*SyncStartResponse, error)
+	SyncWant(context.Context, *SyncWantRequest) (*SyncWantResponse, error)
+	SyncComplete(context.Context, *SyncCompleteRequest) (*SyncCompleteResponse, error)
+	SyncReady(context.Context, *SyncReadyRequest) (*SyncReadyResponse, error)
+	SyncStatus(context.Context, *SyncStatusRequest) (*SyncStatusResponse, error)
+}
+
+type DRPCAgentSocketUnimplementedServer struct{}
+
+func (s *DRPCAgentSocketUnimplementedServer) Ping(context.Context, *PingRequest) (*PingResponse, error) {
+	return nil, drpcerr.WithCode(errors.New("Unimplemented"), drpcerr.Unimplemented)
+}
+
+func (s *DRPCAgentSocketUnimplementedServer) SyncStart(context.Context, *SyncStartRequest) (*SyncStartResponse, error) {
+	return nil, drpcerr.WithCode(errors.New("Unimplemented"), drpcerr.Unimplemented)
+}
+
+func (s *DRPCAgentSocketUnimplementedServer) SyncWant(context.Context, *SyncWantRequest) (*SyncWantResponse, error) {
+	return nil, drpcerr.WithCode(errors.New("Unimplemented"), drpcerr.Unimplemented)
+}
+
+func (s *DRPCAgentSocketUnimplementedServer) SyncComplete(context.Context, *SyncCompleteRequest) (*SyncCompleteResponse, error) {
+	return nil, drpcerr.WithCode(errors.New("Unimplemented"), drpcerr.Unimplemented)
+}
+
+func (s *DRPCAgentSocketUnimplementedServer) SyncReady(context.Context, *SyncReadyRequest) (*SyncReadyResponse, error) {
+	return nil, drpcerr.WithCode(errors.New("Unimplemented"), drpcerr.Unimplemented)
+}
+
+func (s *DRPCAgentSocketUnimplementedServer) SyncStatus(context.Context, *SyncStatusRequest) (*SyncStatusResponse, error) {
+	return nil, drpcerr.WithCode(errors.New("Unimplemented"), drpcerr.Unimplemented)
+}
+
+type DRPCAgentSocketDescription struct{}
+
+func (DRPCAgentSocketDescription) NumMethods() int { return 6 }
+
+func (DRPCAgentSocketDescription) Method(n int) (string, drpc.Encoding, drpc.Receiver, interface{}, bool) {
+	switch n {
+	case 0:
+		return "/coder.agentsocket.v1.AgentSocket/Ping", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{},
+			func(srv interface{}, ctx context.Context, in1, in2 interface{}) (drpc.Message, error) {
+				return srv.(DRPCAgentSocketServer).
+					Ping(
+						ctx,
+						in1.(*PingRequest),
+					)
+			}, DRPCAgentSocketServer.Ping, true
+	case 1:
+		return "/coder.agentsocket.v1.AgentSocket/SyncStart", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{},
+			func(srv interface{}, ctx context.Context, in1, in2 interface{}) (drpc.Message, error) {
+				return srv.(DRPCAgentSocketServer).
+					SyncStart(
+						ctx,
+						in1.(*SyncStartRequest),
+					)
+			}, DRPCAgentSocketServer.SyncStart, true
+	case 2:
+		return "/coder.agentsocket.v1.AgentSocket/SyncWant", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{},
+			func(srv interface{}, ctx context.Context, in1, in2 interface{}) (drpc.Message, error) {
+				return srv.(DRPCAgentSocketServer).
+					SyncWant(
+						ctx,
+						in1.(*SyncWantRequest),
+					)
+			}, DRPCAgentSocketServer.SyncWant, true
+	case 3:
+		return "/coder.agentsocket.v1.AgentSocket/SyncComplete", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{},
+			func(srv interface{}, ctx context.Context, in1, in2 interface{}) (drpc.Message, error) {
+				return srv.(DRPCAgentSocketServer).
+					SyncComplete(
+						ctx,
+						in1.(*SyncCompleteRequest),
+					)
+			}, DRPCAgentSocketServer.SyncComplete, true
+	case 4:
+		return "/coder.agentsocket.v1.AgentSocket/SyncReady", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{},
+			func(srv interface{}, ctx context.Context, in1, in2 interface{}) (drpc.Message, error) {
+				return srv.(DRPCAgentSocketServer).
+					SyncReady(
+						ctx,
+						in1.(*SyncReadyRequest),
+					)
+			}, DRPCAgentSocketServer.SyncReady, true
+	case 5:
+		return "/coder.agentsocket.v1.AgentSocket/SyncStatus", drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{},
+			func(srv interface{}, ctx context.Context, in1, in2 interface{}) (drpc.Message, error) {
+				return srv.(DRPCAgentSocketServer).
+					SyncStatus(
+						ctx,
+						in1.(*SyncStatusRequest),
+					)
+			}, DRPCAgentSocketServer.SyncStatus, true
+	default:
+		return "", nil, nil, nil, false
+	}
+}
+
+func DRPCRegisterAgentSocket(mux drpc.Mux, impl DRPCAgentSocketServer) error {
+	return mux.Register(impl, DRPCAgentSocketDescription{})
+}
+
+type DRPCAgentSocket_PingStream interface {
+	drpc.Stream
+	SendAndClose(*PingResponse) error
+}
+
+type drpcAgentSocket_PingStream struct {
+	drpc.Stream
+}
+
+func (x *drpcAgentSocket_PingStream) SendAndClose(m *PingResponse) error {
+	if err := x.MsgSend(m, drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}); err != nil {
+		return err
+	}
+	return x.CloseSend()
+}
+
+type DRPCAgentSocket_SyncStartStream interface {
+	drpc.Stream
+	SendAndClose(*SyncStartResponse) error
+}
+
+type drpcAgentSocket_SyncStartStream struct {
+	drpc.Stream
+}
+
+func (x *drpcAgentSocket_SyncStartStream) SendAndClose(m *SyncStartResponse) error {
+	if err := x.MsgSend(m, drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}); err != nil {
+		return err
+	}
+	return x.CloseSend()
+}
+
+type DRPCAgentSocket_SyncWantStream interface {
+	drpc.Stream
+	SendAndClose(*SyncWantResponse) error
+}
+
+type drpcAgentSocket_SyncWantStream struct {
+	drpc.Stream
+}
+
+func (x *drpcAgentSocket_SyncWantStream) SendAndClose(m *SyncWantResponse) error {
+	if err := x.MsgSend(m, drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}); err != nil {
+		return err
+	}
+	return x.CloseSend()
+}
+
+type DRPCAgentSocket_SyncCompleteStream interface {
+	drpc.Stream
+	SendAndClose(*SyncCompleteResponse) error
+}
+
+type drpcAgentSocket_SyncCompleteStream struct {
+	drpc.Stream
+}
+
+func (x *drpcAgentSocket_SyncCompleteStream) SendAndClose(m *SyncCompleteResponse) error {
+	if err := x.MsgSend(m, drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}); err != nil {
+		return err
+	}
+	return x.CloseSend()
+}
+
+type DRPCAgentSocket_SyncReadyStream interface {
+	drpc.Stream
+	SendAndClose(*SyncReadyResponse) error
+}
+
+type drpcAgentSocket_SyncReadyStream struct {
+	drpc.Stream
+}
+
+func (x *drpcAgentSocket_SyncReadyStream) SendAndClose(m *SyncReadyResponse) error {
+	if err := x.MsgSend(m, drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}); err != nil {
+		return err
+	}
+	return x.CloseSend()
+}
+
+type DRPCAgentSocket_SyncStatusStream interface {
+	drpc.Stream
+	SendAndClose(*SyncStatusResponse) error
+}
+
+type drpcAgentSocket_SyncStatusStream struct {
+	drpc.Stream
+}
+
+func (x *drpcAgentSocket_SyncStatusStream) SendAndClose(m *SyncStatusResponse) error {
+	if err := x.MsgSend(m, drpcEncoding_File_agent_agentsocket_proto_agentsocket_proto{}); err != nil {
+		return err
+	}
+	return x.CloseSend()
+}

agent/agentsocket/proto/version.go

@@ -0,0 +1,17 @@
+package proto
+
+import "github.com/coder/coder/v2/apiversion"
+
+// Version history:
+//
+// API v1.0:
+//   - Initial release
+//   - Ping
+//   - Sync operations: SyncStart, SyncWant, SyncComplete, SyncWait, SyncStatus
+
+const (
+	CurrentMajor = 1
+	CurrentMinor = 0
+)
+
+var CurrentVersion = apiversion.New(CurrentMajor, CurrentMinor)

agent/agentsocket/server.go

@@ -0,0 +1,138 @@
+package agentsocket
+
+import (
+	"context"
+	"errors"
+	"net"
+	"sync"
+
+	"golang.org/x/xerrors"
+	"storj.io/drpc/drpcmux"
+	"storj.io/drpc/drpcserver"
+
+	"cdr.dev/slog"
+	"github.com/coder/coder/v2/agent/agentsocket/proto"
+	"github.com/coder/coder/v2/agent/unit"
+	"github.com/coder/coder/v2/codersdk/drpcsdk"
+)
+
+// Server provides access to the DRPCAgentSocketService via a Unix domain socket.
+// Do not invoke Server{} directly. Use NewServer() instead.
+type Server struct {
+	logger     slog.Logger
+	path       string
+	drpcServer *drpcserver.Server
+	service    *DRPCAgentSocketService
+
+	mu       sync.Mutex
+	listener net.Listener
+	ctx      context.Context
+	cancel   context.CancelFunc
+	wg       sync.WaitGroup
+}
+
+// NewServer creates a new agent socket server.
+func NewServer(logger slog.Logger, opts ...Option) (*Server, error) {
+	options := &options{}
+	for _, opt := range opts {
+		opt(options)
+	}
+
+	logger = logger.Named("agentsocket-server")
+	server := &Server{
+		logger: logger,
+		path:   options.path,
+		service: &DRPCAgentSocketService{
+			logger:      logger,
+			unitManager: unit.NewManager(),
+		},
+	}
+
+	mux := drpcmux.New()
+	err := proto.DRPCRegisterAgentSocket(mux, server.service)
+	if err != nil {
+		return nil, xerrors.Errorf("failed to register drpc service: %w", err)
+	}
+
+	server.drpcServer = drpcserver.NewWithOptions(mux, drpcserver.Options{
+		Manager: drpcsdk.DefaultDRPCOptions(nil),
+		Log: func(err error) {
+			if errors.Is(err, context.Canceled) ||
+				errors.Is(err, context.DeadlineExceeded) {
+				return
+			}
+			logger.Debug(context.Background(), "drpc server error", slog.Error(err))
+		},
+	})
+
+	listener, err := createSocket(server.path)
+	if err != nil {
+		return nil, xerrors.Errorf("create socket: %w", err)
+	}
+
+	server.listener = listener
+
+	// This context is canceled by server.Close().
+	// canceling it will close all connections.
+	server.ctx, server.cancel = context.WithCancel(context.Background())
+
+	server.logger.Info(server.ctx, "agent socket server started", slog.F("path", server.path))
+
+	server.wg.Add(1)
+	go func() {
+		defer server.wg.Done()
+		server.acceptConnections()
+	}()
+
+	return server, nil
+}
+
+// Close stops the server and cleans up resources.
+func (s *Server) Close() error {
+	s.mu.Lock()
+
+	if s.listener == nil {
+		s.mu.Unlock()
+		return nil
+	}
+
+	s.logger.Info(s.ctx, "stopping agent socket server")
+
+	s.cancel()
+
+	if err := s.listener.Close(); err != nil {
+		s.logger.Warn(s.ctx, "error closing socket listener", slog.Error(err))
+	}
+
+	s.listener = nil
+
+	s.mu.Unlock()
+
+	// Wait for all connections to finish
+	s.wg.Wait()
+
+	if err := cleanupSocket(s.path); err != nil {
+		s.logger.Warn(s.ctx, "error cleaning up socket file", slog.Error(err))
+	}
+
+	s.logger.Info(s.ctx, "agent socket server stopped")
+
+	return nil
+}
+
+func (s *Server) acceptConnections() {
+	// In an edge case, Close() might race with acceptConnections() and set s.listener to nil.
+	// Therefore, we grab a copy of the listener under a lock. We might still get a nil listener,
+	// but then we know close has already run and we can return early.
+	s.mu.Lock()
+	listener := s.listener
+	s.mu.Unlock()
+	if listener == nil {
+		return
+	}
+
+	err := s.drpcServer.Serve(s.ctx, listener)
+	if err != nil {
+		s.logger.Warn(s.ctx, "error serving drpc server", slog.Error(err))
+	}
+}

agent/agentsocket/server_test.go

@@ -0,0 +1,138 @@
+package agentsocket_test
+
+import (
+	"context"
+	"path/filepath"
+	"runtime"
+	"testing"
+
+	"github.com/google/uuid"
+	"github.com/spf13/afero"
+	"github.com/stretchr/testify/require"
+
+	"cdr.dev/slog"
+	"github.com/coder/coder/v2/agent"
+	"github.com/coder/coder/v2/agent/agentsocket"
+	"github.com/coder/coder/v2/agent/agenttest"
+	agentproto "github.com/coder/coder/v2/agent/proto"
+	"github.com/coder/coder/v2/codersdk/agentsdk"
+	"github.com/coder/coder/v2/tailnet"
+	"github.com/coder/coder/v2/tailnet/tailnettest"
+	"github.com/coder/coder/v2/testutil"
+)
+
+func TestServer(t *testing.T) {
+	t.Parallel()
+
+	if runtime.GOOS == "windows" {
+		t.Skip("agentsocket is not supported on Windows")
+	}
+
+	t.Run("StartStop", func(t *testing.T) {
+		t.Parallel()
+
+		socketPath := filepath.Join(t.TempDir(), "test.sock")
+		logger := slog.Make().Leveled(slog.LevelDebug)
+		server, err := agentsocket.NewServer(logger, agentsocket.WithPath(socketPath))
+		require.NoError(t, err)
+		require.NoError(t, server.Close())
+	})
+
+	t.Run("AlreadyStarted", func(t *testing.T) {
+		t.Parallel()
+
+		socketPath := filepath.Join(t.TempDir(), "test.sock")
+		logger := slog.Make().Leveled(slog.LevelDebug)
+		server1, err := agentsocket.NewServer(logger, agentsocket.WithPath(socketPath))
+		require.NoError(t, err)
+		defer server1.Close()
+		_, err = agentsocket.NewServer(logger, agentsocket.WithPath(socketPath))
+		require.ErrorContains(t, err, "create socket")
+	})
+
+	t.Run("AutoSocketPath", func(t *testing.T) {
+		t.Parallel()
+
+		socketPath := filepath.Join(t.TempDir(), "test.sock")
+		logger := slog.Make().Leveled(slog.LevelDebug)
+		server, err := agentsocket.NewServer(logger, agentsocket.WithPath(socketPath))
+		require.NoError(t, err)
+		require.NoError(t, server.Close())
+	})
+}
+
+func TestServerWindowsNotSupported(t *testing.T) {
+	t.Parallel()
+
+	if runtime.GOOS != "windows" {
+		t.Skip("this test only runs on Windows")
+	}
+
+	t.Run("NewServer", func(t *testing.T) {
+		t.Parallel()
+
+		socketPath := filepath.Join(t.TempDir(), "test.sock")
+		logger := slog.Make().Leveled(slog.LevelDebug)
+		_, err := agentsocket.NewServer(logger, agentsocket.WithPath(socketPath))
+		require.ErrorContains(t, err, "agentsocket is not supported on Windows")
+	})
+
+	t.Run("NewClient", func(t *testing.T) {
+		t.Parallel()
+
+		_, err := agentsocket.NewClient(context.Background(), agentsocket.WithPath("test.sock"))
+		require.ErrorContains(t, err, "agentsocket is not supported on Windows")
+	})
+}
+
+func TestAgentInitializesOnWindowsWithoutSocketServer(t *testing.T) {
+	t.Parallel()
+
+	if runtime.GOOS != "windows" {
+		t.Skip("this test only runs on Windows")
+	}
+
+	ctx := testutil.Context(t, testutil.WaitShort)
+	logger := testutil.Logger(t).Named("agent")
+
+	derpMap, _ := tailnettest.RunDERPAndSTUN(t)
+
+	coordinator := tailnet.NewCoordinator(logger)
+	t.Cleanup(func() {
+		_ = coordinator.Close()
+	})
+
+	statsCh := make(chan *agentproto.Stats, 50)
+	agentID := uuid.New()
+	manifest := agentsdk.Manifest{
+		AgentID:       agentID,
+		AgentName:     "test-agent",
+		WorkspaceName: "test-workspace",
+		OwnerName:     "test-user",
+		WorkspaceID:   uuid.New(),
+		DERPMap:       derpMap,
+	}
+
+	client := agenttest.NewClient(t, logger.Named("agenttest"), agentID, manifest, statsCh, coordinator)
+	t.Cleanup(client.Close)
+
+	options := agent.Options{
+		Client:                 client,
+		Filesystem:             afero.NewMemMapFs(),
+		Logger:                 logger.Named("agent"),
+		ReconnectingPTYTimeout: testutil.WaitShort,
+		EnvironmentVariables:   map[string]string{},
+		SocketPath:             "",
+	}
+
+	agnt := agent.New(options)
+	t.Cleanup(func() {
+		_ = agnt.Close()
+	})
+
+	startup := testutil.TryReceive(ctx, t, client.GetStartup())
+	require.NotNil(t, startup, "agent should send startup message")
+
+	err := agnt.Close()
+	require.NoError(t, err, "agent should close cleanly")
+}

agent/agentsocket/service.go

@@ -0,0 +1,152 @@
+package agentsocket
+
+import (
+	"context"
+	"errors"
+
+	"golang.org/x/xerrors"
+
+	"cdr.dev/slog"
+	"github.com/coder/coder/v2/agent/agentsocket/proto"
+	"github.com/coder/coder/v2/agent/unit"
+)
+
+var _ proto.DRPCAgentSocketServer = (*DRPCAgentSocketService)(nil)
+
+var ErrUnitManagerNotAvailable = xerrors.New("unit manager not available")
+
+// DRPCAgentSocketService implements the DRPC agent socket service.
+type DRPCAgentSocketService struct {
+	unitManager *unit.Manager
+	logger      slog.Logger
+}
+
+// Ping responds to a ping request to check if the service is alive.
+func (*DRPCAgentSocketService) Ping(_ context.Context, _ *proto.PingRequest) (*proto.PingResponse, error) {
+	return &proto.PingResponse{}, nil
+}
+
+// SyncStart starts a unit in the dependency graph.
+func (s *DRPCAgentSocketService) SyncStart(_ context.Context, req *proto.SyncStartRequest) (*proto.SyncStartResponse, error) {
+	if s.unitManager == nil {
+		return nil, xerrors.Errorf("SyncStart: %w", ErrUnitManagerNotAvailable)
+	}
+
+	unitID := unit.ID(req.Unit)
+
+	if err := s.unitManager.Register(unitID); err != nil {
+		if !errors.Is(err, unit.ErrUnitAlreadyRegistered) {
+			return nil, xerrors.Errorf("SyncStart: %w", err)
+		}
+	}
+
+	isReady, err := s.unitManager.IsReady(unitID)
+	if err != nil {
+		return nil, xerrors.Errorf("cannot check readiness: %w", err)
+	}
+	if !isReady {
+		return nil, xerrors.Errorf("cannot start unit %q: unit not ready", req.Unit)
+	}
+
+	err = s.unitManager.UpdateStatus(unitID, unit.StatusStarted)
+	if err != nil {
+		return nil, xerrors.Errorf("cannot start unit %q: %w", req.Unit, err)
+	}
+
+	return &proto.SyncStartResponse{}, nil
+}
+
+// SyncWant declares a dependency between units.
+func (s *DRPCAgentSocketService) SyncWant(_ context.Context, req *proto.SyncWantRequest) (*proto.SyncWantResponse, error) {
+	if s.unitManager == nil {
+		return nil, xerrors.Errorf("cannot add dependency: %w", ErrUnitManagerNotAvailable)
+	}
+
+	unitID := unit.ID(req.Unit)
+	dependsOnID := unit.ID(req.DependsOn)
+
+	if err := s.unitManager.Register(unitID); err != nil && !errors.Is(err, unit.ErrUnitAlreadyRegistered) {
+		return nil, xerrors.Errorf("cannot add dependency: %w", err)
+	}
+
+	if err := s.unitManager.AddDependency(unitID, dependsOnID, unit.StatusComplete); err != nil {
+		return nil, xerrors.Errorf("cannot add dependency: %w", err)
+	}
+
+	return &proto.SyncWantResponse{}, nil
+}
+
+// SyncComplete marks a unit as complete in the dependency graph.
+func (s *DRPCAgentSocketService) SyncComplete(_ context.Context, req *proto.SyncCompleteRequest) (*proto.SyncCompleteResponse, error) {
+	if s.unitManager == nil {
+		return nil, xerrors.Errorf("cannot complete unit: %w", ErrUnitManagerNotAvailable)
+	}
+
+	unitID := unit.ID(req.Unit)
+
+	if err := s.unitManager.UpdateStatus(unitID, unit.StatusComplete); err != nil {
+		return nil, xerrors.Errorf("cannot complete unit %q: %w", req.Unit, err)
+	}
+
+	return &proto.SyncCompleteResponse{}, nil
+}
+
+// SyncReady checks whether a unit is ready to be started. That is, all dependencies are satisfied.
+func (s *DRPCAgentSocketService) SyncReady(_ context.Context, req *proto.SyncReadyRequest) (*proto.SyncReadyResponse, error) {
+	if s.unitManager == nil {
+		return nil, xerrors.Errorf("cannot check readiness: %w", ErrUnitManagerNotAvailable)
+	}
+
+	unitID := unit.ID(req.Unit)
+	isReady, err := s.unitManager.IsReady(unitID)
+	if err != nil {
+		return nil, xerrors.Errorf("cannot check readiness: %w", err)
+	}
+
+	return &proto.SyncReadyResponse{
+		Ready: isReady,
+	}, nil
+}
+
+// SyncStatus gets the status of a unit and lists its dependencies.
+func (s *DRPCAgentSocketService) SyncStatus(_ context.Context, req *proto.SyncStatusRequest) (*proto.SyncStatusResponse, error) {
+	if s.unitManager == nil {
+		return nil, xerrors.Errorf("cannot get status for unit %q: %w", req.Unit, ErrUnitManagerNotAvailable)
+	}
+
+	unitID := unit.ID(req.Unit)
+
+	isReady, err := s.unitManager.IsReady(unitID)
+	if err != nil {
+		return nil, xerrors.Errorf("cannot check readiness: %w", err)
+	}
+
+	dependencies, err := s.unitManager.GetAllDependencies(unitID)
+	switch {
+	case errors.Is(err, unit.ErrUnitNotFound):
+		dependencies = []unit.Dependency{}
+	case err != nil:
+		return nil, xerrors.Errorf("cannot get dependencies: %w", err)
+	}
+
+	var depInfos []*proto.DependencyInfo
+	for _, dep := range dependencies {
+		depInfos = append(depInfos, &proto.DependencyInfo{
+			Unit:           string(dep.Unit),
+			DependsOn:      string(dep.DependsOn),
+			RequiredStatus: string(dep.RequiredStatus),
+			CurrentStatus:  string(dep.CurrentStatus),
+			IsSatisfied:    dep.IsSatisfied,
+		})
+	}
+
+	u, err := s.unitManager.Unit(unitID)
+	if err != nil {
+		return nil, xerrors.Errorf("cannot get status for unit %q: %w", req.Unit, err)
+	}
+	return &proto.SyncStatusResponse{
+		Status:       string(u.Status()),
+		IsReady:      isReady,
+		Dependencies: depInfos,
+	}, nil
+}

agent/agentsocket/service_test.go

@@ -0,0 +1,389 @@
+package agentsocket_test
+
+import (
+	"context"
+	"crypto/sha256"
+	"encoding/hex"
+	"fmt"
+	"os"
+	"path/filepath"
+	"runtime"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+
+	"cdr.dev/slog"
+	"github.com/coder/coder/v2/agent/agentsocket"
+	"github.com/coder/coder/v2/agent/unit"
+	"github.com/coder/coder/v2/testutil"
+)
+
+// tempDirUnixSocket returns a temporary directory that can safely hold unix
+// sockets (probably).
+//
+// During tests on darwin we hit the max path length limit for unix sockets
+// pretty easily in the default location, so this function uses /tmp instead to
+// get shorter paths. To keep paths short, we use a hash of the test name
+// instead of the full test name.
+func tempDirUnixSocket(t *testing.T) string {
+	t.Helper()
+	if runtime.GOOS == "darwin" {
+		// Use a short hash of the test name to keep the path under 104 chars
+		hash := sha256.Sum256([]byte(t.Name()))
+		hashStr := hex.EncodeToString(hash[:])[:8] // Use first 8 chars of hash
+		dir, err := os.MkdirTemp("/tmp", fmt.Sprintf("c-%s-", hashStr))
+		require.NoError(t, err, "create temp dir for unix socket test")
+		t.Cleanup(func() {
+			err := os.RemoveAll(dir)
+			assert.NoError(t, err, "remove temp dir", dir)
+		})
+		return dir
+	}
+	return t.TempDir()
+}
+
+// newSocketClient creates a DRPC client connected to the Unix socket at the given path.
+func newSocketClient(ctx context.Context, t *testing.T, socketPath string) *agentsocket.Client {
+	t.Helper()
+
+	client, err := agentsocket.NewClient(ctx, agentsocket.WithPath(socketPath))
+	t.Cleanup(func() {
+		_ = client.Close()
+	})
+	require.NoError(t, err)
+
+	return client
+}
+
+func TestDRPCAgentSocketService(t *testing.T) {
+	t.Parallel()
+
+	if runtime.GOOS == "windows" {
+		t.Skip("agentsocket is not supported on Windows")
+	}
+
+	t.Run("Ping", func(t *testing.T) {
+		t.Parallel()
+
+		socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+		ctx := testutil.Context(t, testutil.WaitShort)
+		server, err := agentsocket.NewServer(
+			slog.Make().Leveled(slog.LevelDebug),
+			agentsocket.WithPath(socketPath),
+		)
+		require.NoError(t, err)
+		defer server.Close()
+
+		client := newSocketClient(ctx, t, socketPath)
+
+		err = client.Ping(ctx)
+		require.NoError(t, err)
+	})
+
+	t.Run("SyncStart", func(t *testing.T) {
+		t.Parallel()
+
+		t.Run("NewUnit", func(t *testing.T) {
+			t.Parallel()
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			err = client.SyncStart(ctx, "test-unit")
+			require.NoError(t, err)
+
+			status, err := client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusStarted, status.Status)
+		})
+
+		t.Run("UnitAlreadyStarted", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			// First Start
+			err = client.SyncStart(ctx, "test-unit")
+			require.NoError(t, err)
+			status, err := client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusStarted, status.Status)
+
+			// Second Start
+			err = client.SyncStart(ctx, "test-unit")
+			require.ErrorContains(t, err, unit.ErrSameStatusAlreadySet.Error())
+
+			status, err = client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusStarted, status.Status)
+		})
+
+		t.Run("UnitAlreadyCompleted", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			// First start
+			err = client.SyncStart(ctx, "test-unit")
+			require.NoError(t, err)
+
+			status, err := client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusStarted, status.Status)
+
+			// Complete the unit
+			err = client.SyncComplete(ctx, "test-unit")
+			require.NoError(t, err)
+
+			status, err = client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusComplete, status.Status)
+
+			// Second start
+			err = client.SyncStart(ctx, "test-unit")
+			require.NoError(t, err)
+
+			status, err = client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusStarted, status.Status)
+		})
+
+		t.Run("UnitNotReady", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			err = client.SyncWant(ctx, "test-unit", "dependency-unit")
+			require.NoError(t, err)
+
+			err = client.SyncStart(ctx, "test-unit")
+			require.ErrorContains(t, err, "unit not ready")
+
+			status, err := client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusPending, status.Status)
+			require.False(t, status.IsReady)
+		})
+	})
+
+	t.Run("SyncWant", func(t *testing.T) {
+		t.Parallel()
+
+		t.Run("NewUnits", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			// If dependency units are not registered, they are registered automatically
+			err = client.SyncWant(ctx, "test-unit", "dependency-unit")
+			require.NoError(t, err)
+
+			status, err := client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Len(t, status.Dependencies, 1)
+			require.Equal(t, unit.ID("dependency-unit"), status.Dependencies[0].DependsOn)
+			require.Equal(t, unit.StatusComplete, status.Dependencies[0].RequiredStatus)
+		})
+
+		t.Run("DependencyAlreadyRegistered", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			// Start the dependency unit
+			err = client.SyncStart(ctx, "dependency-unit")
+			require.NoError(t, err)
+
+			status, err := client.SyncStatus(ctx, "dependency-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusStarted, status.Status)
+
+			// Add the dependency after the dependency unit has already started
+			err = client.SyncWant(ctx, "test-unit", "dependency-unit")
+
+			// Dependencies can be added even if the dependency unit has already started
+			require.NoError(t, err)
+
+			// The dependency is now reflected in the test unit's status
+			status, err = client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.ID("dependency-unit"), status.Dependencies[0].DependsOn)
+			require.Equal(t, unit.StatusComplete, status.Dependencies[0].RequiredStatus)
+		})
+
+		t.Run("DependencyAddedAfterDependentStarted", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			// Start the dependent unit
+			err = client.SyncStart(ctx, "test-unit")
+			require.NoError(t, err)
+
+			status, err := client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.StatusStarted, status.Status)
+
+			// Add the dependency after the dependency unit has already started
+			err = client.SyncWant(ctx, "test-unit", "dependency-unit")
+
+			// Dependencies can be added even if the dependent unit has already started.
+			// The dependency applies the next time a unit is started. The current status is not updated.
+			// This is to allow flexible dependency management. It does mean that users of this API should
+			// take care to add dependencies before they start their dependent units.
+			require.NoError(t, err)
+
+			// The dependency is now reflected in the test unit's status
+			status, err = client.SyncStatus(ctx, "test-unit")
+			require.NoError(t, err)
+			require.Equal(t, unit.ID("dependency-unit"), status.Dependencies[0].DependsOn)
+			require.Equal(t, unit.StatusComplete, status.Dependencies[0].RequiredStatus)
+		})
+	})
+
+	t.Run("SyncReady", func(t *testing.T) {
+		t.Parallel()
+
+		t.Run("UnregisteredUnit", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			ready, err := client.SyncReady(ctx, "unregistered-unit")
+			require.NoError(t, err)
+			require.True(t, ready)
+		})
+
+		t.Run("UnitNotReady", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			// Register a unit with an unsatisfied dependency
+			err = client.SyncWant(ctx, "test-unit", "dependency-unit")
+			require.NoError(t, err)
+
+			// Check readiness - should be false because dependency is not satisfied
+			ready, err := client.SyncReady(ctx, "test-unit")
+			require.NoError(t, err)
+			require.False(t, ready)
+		})
+
+		t.Run("UnitReady", func(t *testing.T) {
+			t.Parallel()
+
+			socketPath := filepath.Join(tempDirUnixSocket(t), "test.sock")
+			ctx := testutil.Context(t, testutil.WaitShort)
+			server, err := agentsocket.NewServer(
+				slog.Make().Leveled(slog.LevelDebug),
+				agentsocket.WithPath(socketPath),
+			)
+			require.NoError(t, err)
+			defer server.Close()
+
+			client := newSocketClient(ctx, t, socketPath)
+
+			// Register a unit with no dependencies - should be ready immediately
+			err = client.SyncStart(ctx, "test-unit")
+			require.NoError(t, err)
+
+			// Check readiness - should be true
+			ready, err := client.SyncReady(ctx, "test-unit")
+			require.NoError(t, err)
+			require.True(t, ready)
+
+			// Also test a unit with satisfied dependencies
+			err = client.SyncWant(ctx, "dependent-unit", "test-unit")
+			require.NoError(t, err)
+
+			// Complete the dependency
+			err = client.SyncComplete(ctx, "test-unit")
+			require.NoError(t, err)
+
+			// Now dependent-unit should be ready
+			ready, err = client.SyncReady(ctx, "dependent-unit")
+			require.NoError(t, err)
+			require.True(t, ready)
+		})
+	})
+}

agent/agentsocket/socket_unix.go

@@ -0,0 +1,73 @@
+//go:build !windows
+
+package agentsocket
+
+import (
+	"context"
+	"net"
+	"os"
+	"path/filepath"
+	"time"
+
+	"golang.org/x/xerrors"
+)
+
+const defaultSocketPath = "/tmp/coder-agent.sock"
+
+func createSocket(path string) (net.Listener, error) {
+	if path == "" {
+		path = defaultSocketPath
+	}
+
+	if !isSocketAvailable(path) {
+		return nil, xerrors.Errorf("socket path %s is not available", path)
+	}
+
+	if err := os.Remove(path); err != nil && !os.IsNotExist(err) {
+		return nil, xerrors.Errorf("remove existing socket: %w", err)
+	}
+
+	parentDir := filepath.Dir(path)
+	if err := os.MkdirAll(parentDir, 0o700); err != nil {
+		return nil, xerrors.Errorf("create socket directory: %w", err)
+	}
+
+	listener, err := net.Listen("unix", path)
+	if err != nil {
+		return nil, xerrors.Errorf("listen on unix socket: %w", err)
+	}
+
+	if err := os.Chmod(path, 0o600); err != nil {
+		_ = listener.Close()
+		return nil, xerrors.Errorf("set socket permissions: %w", err)
+	}
+	return listener, nil
+}
+
+func cleanupSocket(path string) error {
+	return os.Remove(path)
+}
+
+func isSocketAvailable(path string) bool {
+	if _, err := os.Stat(path); os.IsNotExist(err) {
+		return true
+	}
+
+	// Try to connect to see if it's actually listening.
+	dialer := net.Dialer{Timeout: 10 * time.Second}
+	conn, err := dialer.Dial("unix", path)
+	if err != nil {
+		return true
+	}
+	_ = conn.Close()
+	return false
+}
+
+func dialSocket(ctx context.Context, path string) (net.Conn, error) {
+	if path == "" {
+		path = defaultSocketPath
+	}
+
+	dialer := net.Dialer{}
+	return dialer.DialContext(ctx, "unix", path)
+}

agent/agentsocket/socket_windows.go

@@ -0,0 +1,22 @@
+//go:build windows
+
+package agentsocket
+
+import (
+	"context"
+	"net"
+
+	"golang.org/x/xerrors"
+)
+
+func createSocket(_ string) (net.Listener, error) {
+	return nil, xerrors.New("agentsocket is not supported on Windows")
+}
+
+func cleanupSocket(_ string) error {
+	return nil
+}
+
+func dialSocket(_ context.Context, _ string) (net.Conn, error) {
+	return nil, xerrors.New("agentsocket is not supported on Windows")
+}

agent/agentssh/agentssh.go

@@ -46,6 +46,8 @@ const (
 	// MagicProcessCmdlineJetBrains is a string in a process's command line that
 	// uniquely identifies it as JetBrains software.
 	MagicProcessCmdlineJetBrains = "idea.vendor.name=JetBrains"
+	MagicProcessCmdlineToolbox   = "com.jetbrains.toolbox"
+	MagicProcessCmdlineGateway   = "remote-dev-server"
 
 	// BlockedFileTransferErrorCode indicates that SSH server restricted the raw command from performing
 	// the file transfer.
@@ -117,6 +119,10 @@ type Config struct {
 	// Note that this is different from the devcontainers feature, which uses
 	// subagents.
 	ExperimentalContainers bool
+	// X11Net allows overriding the networking implementation used for X11
+	// forwarding listeners. When nil, a default implementation backed by the
+	// standard library networking package is used.
+	X11Net X11Network
 }
 
 type Server struct {
@@ -131,9 +137,10 @@ type Server struct {
 	// a lock on mu but protected by closing.
 	wg sync.WaitGroup
 
-	Execer agentexec.Execer
-	logger slog.Logger
-	srv    *ssh.Server
+	Execer       agentexec.Execer
+	logger       slog.Logger
+	srv          *ssh.Server
+	x11Forwarder *x11Forwarder
 
 	config *Config
 
@@ -190,6 +197,20 @@ func NewServer(ctx context.Context, logger slog.Logger, prometheusRegistry *prom
 		config: config,
 
 		metrics: metrics,
+		x11Forwarder: &x11Forwarder{
+			logger:           logger,
+			x11HandlerErrors: metrics.x11HandlerErrors,
+			fs:               fs,
+			displayOffset:    *config.X11DisplayOffset,
+			sessions:         make(map[*x11Session]struct{}),
+			connections:      make(map[net.Conn]struct{}),
+			network: func() X11Network {
+				if config.X11Net != nil {
+					return config.X11Net
+				}
+				return osNet{}
+			}(),
+		},
 	}
 
 	srv := &ssh.Server{
@@ -370,10 +391,19 @@ func (s *Server) sessionHandler(session ssh.Session) {
 	env := session.Environ()
 	magicType, magicTypeRaw, env := extractMagicSessionType(env)
 
+	// It's not safe to assume RemoteAddr() returns a non-nil value. slog.F usage is fine because it correctly
+	// handles nil.
+	// c.f. https://github.com/coder/internal/issues/1143
+	remoteAddr := session.RemoteAddr()
+	remoteAddrString := ""
+	if remoteAddr != nil {
+		remoteAddrString = remoteAddr.String()
+	}
+
 	if !s.trackSession(session, true) {
 		reason := "unable to accept new session, server is closing"
 		// Report connection attempt even if we couldn't accept it.
-		disconnected := s.config.ReportConnection(id, magicType, session.RemoteAddr().String())
+		disconnected := s.config.ReportConnection(id, magicType, remoteAddrString)
 		defer disconnected(1, reason)
 
 		logger.Info(ctx, reason)
@@ -408,7 +438,7 @@ func (s *Server) sessionHandler(session ssh.Session) {
 		scr := &sessionCloseTracker{Session: session}
 		session = scr
 
-		disconnected := s.config.ReportConnection(id, magicType, session.RemoteAddr().String())
+		disconnected := s.config.ReportConnection(id, magicType, remoteAddrString)
 		defer func() {
 			disconnected(scr.exitCode(), reason)
 		}()
@@ -457,7 +487,7 @@ func (s *Server) sessionHandler(session ssh.Session) {
 
 	x11, hasX11 := session.X11()
 	if hasX11 {
-		display, handled := s.x11Handler(ctx, x11)
+		display, handled := s.x11Forwarder.x11Handler(ctx, session)
 		if !handled {
 			logger.Error(ctx, "x11 handler failed")
 			closeCause("x11 handler failed")
@@ -590,7 +620,9 @@ func (s *Server) startNonPTYSession(logger slog.Logger, session ssh.Session, mag
 	// and SSH server close may be delayed.
 	cmd.SysProcAttr = cmdSysProcAttr()
 
-	// to match OpenSSH, we don't actually tear a non-TTY command down, even if the session ends.
+	// to match OpenSSH, we don't actually tear a non-TTY command down, even if the session ends. OpenSSH closes the
+	// pipes to the process when the session ends; which is what happens here since we wire the command up to the
+	// session for I/O.
 	// c.f. https://github.com/coder/coder/issues/18519#issuecomment-3019118271
 	cmd.Cancel = nil
 
@@ -1154,6 +1186,9 @@ func (s *Server) Close() error {
 
 	s.mu.Unlock()
 
+	s.logger.Debug(ctx, "closing X11 forwarding")
+	_ = s.x11Forwarder.Close()
+
 	s.logger.Debug(ctx, "waiting for all goroutines to exit")
 	s.wg.Wait() // Wait for all goroutines to exit.
 

agent/agentssh/agentssh_test.go

@@ -8,7 +8,9 @@ import (
 	"context"
 	"fmt"
 	"net"
+	"os"
 	"os/user"
+	"path/filepath"
 	"runtime"
 	"strings"
 	"sync"
@@ -403,6 +405,92 @@ func TestNewServer_Signal(t *testing.T) {
 	})
 }
 
+func TestSSHServer_ClosesStdin(t *testing.T) {
+	t.Parallel()
+	if runtime.GOOS == "windows" {
+		t.Skip("bash doesn't exist on Windows")
+	}
+
+	ctx := testutil.Context(t, testutil.WaitMedium)
+	logger := testutil.Logger(t)
+	s, err := agentssh.NewServer(ctx, logger.Named("ssh-server"), prometheus.NewRegistry(), afero.NewMemMapFs(), agentexec.DefaultExecer, nil)
+	require.NoError(t, err)
+	logger = logger.Named("test")
+	defer s.Close()
+	err = s.UpdateHostSigner(42)
+	assert.NoError(t, err)
+
+	ln, err := net.Listen("tcp", "127.0.0.1:0")
+	require.NoError(t, err)
+
+	done := make(chan struct{})
+	go func() {
+		defer close(done)
+		err := s.Serve(ln)
+		assert.Error(t, err) // Server is closed.
+	}()
+	defer func() {
+		err := s.Close()
+		require.NoError(t, err)
+		<-done
+	}()
+
+	c := sshClient(t, ln.Addr().String())
+
+	sess, err := c.NewSession()
+	require.NoError(t, err)
+	stdout, err := sess.StdoutPipe()
+	require.NoError(t, err)
+	stdin, err := sess.StdinPipe()
+	require.NoError(t, err)
+	defer stdin.Close()
+
+	dir := t.TempDir()
+	err = os.MkdirAll(dir, 0o755)
+	require.NoError(t, err)
+	filePath := filepath.Join(dir, "result.txt")
+
+	// the shell command `read` will block until data is written to stdin, or closed. It will return
+	// exit code 1 if it hits EOF, which is what we want to test.
+	cmdErrCh := make(chan error, 1)
+	go func() {
+		cmdErrCh <- sess.Start(fmt.Sprintf(`echo started; echo "read exit code: $(read && echo 0 || echo 1)" > %s`, filePath))
+	}()
+
+	cmdErr := testutil.RequireReceive(ctx, t, cmdErrCh)
+	require.NoError(t, cmdErr)
+
+	readCh := make(chan error, 1)
+	go func() {
+		buf := make([]byte, 8)
+		_, err := stdout.Read(buf)
+		assert.Equal(t, "started\n", string(buf))
+		readCh <- err
+	}()
+	err = testutil.RequireReceive(ctx, t, readCh)
+	require.NoError(t, err)
+
+	err = sess.Close()
+	require.NoError(t, err)
+
+	var content []byte
+	expected := []byte("read exit code: 1\n")
+	testutil.Eventually(ctx, t, func(_ context.Context) bool {
+		content, err = os.ReadFile(filePath)
+		if err != nil {
+			logger.Debug(ctx, "failed to read file; will retry", slog.Error(err))
+			return false
+		}
+		if len(content) != len(expected) {
+			logger.Debug(ctx, "file is partially written", slog.F("content", content))
+			return false
+		}
+		return true
+	}, testutil.IntervalFast)
+	require.NoError(t, err)
+	require.Equal(t, string(expected), string(content))
+}
+
 func sshClient(t *testing.T, addr string) *ssh.Client {
 	conn, err := net.Dial("tcp", addr)
 	require.NoError(t, err)

agent/agentssh/jetbrainstrack.go

@@ -53,7 +53,7 @@ func NewJetbrainsChannelWatcher(ctx ssh.Context, logger slog.Logger, reportConne
 
 	// If this is not JetBrains, then we do not need to do anything special.  We
 	// attempt to match on something that appears unique to JetBrains software.
-	if !strings.Contains(strings.ToLower(cmdline), strings.ToLower(MagicProcessCmdlineJetBrains)) {
+	if !isJetbrainsProcess(cmdline) {
 		return newChannel
 	}
 
@@ -104,3 +104,18 @@ func (c *ChannelOnClose) Close() error {
 	c.once.Do(c.done)
 	return c.Channel.Close()
 }
+
+func isJetbrainsProcess(cmdline string) bool {
+	opts := []string{
+		MagicProcessCmdlineJetBrains,
+		MagicProcessCmdlineToolbox,
+		MagicProcessCmdlineGateway,
+	}
+
+	for _, opt := range opts {
+		if strings.Contains(strings.ToLower(cmdline), strings.ToLower(opt)) {
+			return true
+		}
+	}
+	return false
+}

agent/agentssh/x11.go

@@ -7,15 +7,16 @@ import (
 	"errors"
 	"fmt"
 	"io"
-	"math"
 	"net"
 	"os"
 	"path/filepath"
 	"strconv"
+	"sync"
 	"time"
 
 	"github.com/gliderlabs/ssh"
 	"github.com/gofrs/flock"
+	"github.com/prometheus/client_golang/prometheus"
 	"github.com/spf13/afero"
 	gossh "golang.org/x/crypto/ssh"
 	"golang.org/x/xerrors"
@@ -29,8 +30,51 @@ const (
 	X11StartPort = 6000
 	// X11DefaultDisplayOffset is the default offset for X11 forwarding.
 	X11DefaultDisplayOffset = 10
+	X11MaxDisplays          = 200
+	// X11MaxPort is the highest port we will ever use for X11 forwarding. This limits the total number of TCP sockets
+	// we will create. It seems more useful to have a maximum port number than a direct limit on sockets with no max
+	// port because we'd like to be able to tell users the exact range of ports the Agent might use.
+	X11MaxPort = X11StartPort + X11MaxDisplays
 )
 
+// X11Network abstracts the creation of network listeners for X11 forwarding.
+// It is intended mainly for testing; production code uses the default
+// implementation backed by the operating system networking stack.
+type X11Network interface {
+	Listen(network, address string) (net.Listener, error)
+}
+
+// osNet is the default X11Network implementation that uses the standard
+// library network stack.
+type osNet struct{}
+
+func (osNet) Listen(network, address string) (net.Listener, error) {
+	return net.Listen(network, address)
+}
+
+type x11Forwarder struct {
+	logger           slog.Logger
+	x11HandlerErrors *prometheus.CounterVec
+	fs               afero.Fs
+	displayOffset    int
+
+	// network creates X11 listener sockets. Defaults to osNet{}.
+	network X11Network
+
+	mu          sync.Mutex
+	sessions    map[*x11Session]struct{}
+	connections map[net.Conn]struct{}
+	closing     bool
+	wg          sync.WaitGroup
+}
+
+type x11Session struct {
+	session  ssh.Session
+	display  int
+	listener net.Listener
+	usedAt   time.Time
+}
+
 // x11Callback is called when the client requests X11 forwarding.
 func (*Server) x11Callback(_ ssh.Context, _ ssh.X11) bool {
 	// Always allow.
@@ -39,115 +83,243 @@ func (*Server) x11Callback(_ ssh.Context, _ ssh.X11) bool {
 
 // x11Handler is called when a session has requested X11 forwarding.
 // It listens for X11 connections and forwards them to the client.
-func (s *Server) x11Handler(ctx ssh.Context, x11 ssh.X11) (displayNumber int, handled bool) {
-	serverConn, valid := ctx.Value(ssh.ContextKeyConn).(*gossh.ServerConn)
-	if !valid {
-		s.logger.Warn(ctx, "failed to get server connection")
+func (x *x11Forwarder) x11Handler(sshCtx ssh.Context, sshSession ssh.Session) (displayNumber int, handled bool) {
+	x11, hasX11 := sshSession.X11()
+	if !hasX11 {
 		return -1, false
 	}
+	serverConn, valid := sshCtx.Value(ssh.ContextKeyConn).(*gossh.ServerConn)
+	if !valid {
+		x.logger.Warn(sshCtx, "failed to get server connection")
+		return -1, false
+	}
+	ctx := slog.With(sshCtx, slog.F("session_id", fmt.Sprintf("%x", serverConn.SessionID())))
 
 	hostname, err := os.Hostname()
 	if err != nil {
-		s.logger.Warn(ctx, "failed to get hostname", slog.Error(err))
-		s.metrics.x11HandlerErrors.WithLabelValues("hostname").Add(1)
+		x.logger.Warn(ctx, "failed to get hostname", slog.Error(err))
+		x.x11HandlerErrors.WithLabelValues("hostname").Add(1)
 		return -1, false
 	}
 
-	ln, display, err := createX11Listener(ctx, *s.config.X11DisplayOffset)
+	x11session, err := x.createX11Session(ctx, sshSession)
 	if err != nil {
-		s.logger.Warn(ctx, "failed to create X11 listener", slog.Error(err))
-		s.metrics.x11HandlerErrors.WithLabelValues("listen").Add(1)
+		x.logger.Warn(ctx, "failed to create X11 listener", slog.Error(err))
+		x.x11HandlerErrors.WithLabelValues("listen").Add(1)
 		return -1, false
 	}
-	s.trackListener(ln, true)
 	defer func() {
 		if !handled {
-			s.trackListener(ln, false)
-			_ = ln.Close()
+			x.closeAndRemoveSession(x11session)
 		}
 	}()
 
-	err = addXauthEntry(ctx, s.fs, hostname, strconv.Itoa(display), x11.AuthProtocol, x11.AuthCookie)
+	err = addXauthEntry(ctx, x.fs, hostname, strconv.Itoa(x11session.display), x11.AuthProtocol, x11.AuthCookie)
 	if err != nil {
-		s.logger.Warn(ctx, "failed to add Xauthority entry", slog.Error(err))
-		s.metrics.x11HandlerErrors.WithLabelValues("xauthority").Add(1)
+		x.logger.Warn(ctx, "failed to add Xauthority entry", slog.Error(err))
+		x.x11HandlerErrors.WithLabelValues("xauthority").Add(1)
 		return -1, false
 	}
 
+	// clean up the X11 session if the SSH session completes.
 	go func() {
-		// Don't leave the listener open after the session is gone.
 		<-ctx.Done()
-		_ = ln.Close()
+		x.closeAndRemoveSession(x11session)
 	}()
 
-	go func() {
-		defer ln.Close()
-		defer s.trackListener(ln, false)
+	go x.listenForConnections(ctx, x11session, serverConn, x11)
+	x.logger.Debug(ctx, "X11 forwarding started", slog.F("display", x11session.display))
 
-		for {
-			conn, err := ln.Accept()
-			if err != nil {
-				if errors.Is(err, net.ErrClosed) {
-					return
-				}
-				s.logger.Warn(ctx, "failed to accept X11 connection", slog.Error(err))
+	return x11session.display, true
+}
+
+func (x *x11Forwarder) trackGoroutine() (closing bool, done func()) {
+	x.mu.Lock()
+	defer x.mu.Unlock()
+	if !x.closing {
+		x.wg.Add(1)
+		return false, func() { x.wg.Done() }
+	}
+	return true, func() {}
+}
+
+func (x *x11Forwarder) listenForConnections(
+	ctx context.Context, session *x11Session, serverConn *gossh.ServerConn, x11 ssh.X11,
+) {
+	defer x.closeAndRemoveSession(session)
+	if closing, done := x.trackGoroutine(); closing {
+		return
+	} else { // nolint: revive
+		defer done()
+	}
+
+	for {
+		conn, err := session.listener.Accept()
+		if err != nil {
+			if errors.Is(err, net.ErrClosed) {
 				return
 			}
-			if x11.SingleConnection {
-				s.logger.Debug(ctx, "single connection requested, closing X11 listener")
-				_ = ln.Close()
-			}
-
-			tcpConn, ok := conn.(*net.TCPConn)
-			if !ok {
-				s.logger.Warn(ctx, fmt.Sprintf("failed to cast connection to TCPConn. got: %T", conn))
-				_ = conn.Close()
-				continue
-			}
-			tcpAddr, ok := tcpConn.LocalAddr().(*net.TCPAddr)
-			if !ok {
-				s.logger.Warn(ctx, fmt.Sprintf("failed to cast local address to TCPAddr. got: %T", tcpConn.LocalAddr()))
-				_ = conn.Close()
-				continue
-			}
-
-			channel, reqs, err := serverConn.OpenChannel("x11", gossh.Marshal(struct {
-				OriginatorAddress string
-				OriginatorPort    uint32
-			}{
-				OriginatorAddress: tcpAddr.IP.String(),
-				// #nosec G115 - Safe conversion as TCP port numbers are within uint32 range (0-65535)
-				OriginatorPort: uint32(tcpAddr.Port),
-			}))
-			if err != nil {
-				s.logger.Warn(ctx, "failed to open X11 channel", slog.Error(err))
-				_ = conn.Close()
-				continue
-			}
-			go gossh.DiscardRequests(reqs)
-
-			if !s.trackConn(ln, conn, true) {
-				s.logger.Warn(ctx, "failed to track X11 connection")
-				_ = conn.Close()
-				continue
-			}
-			go func() {
-				defer s.trackConn(ln, conn, false)
-				Bicopy(ctx, conn, channel)
-			}()
+			x.logger.Warn(ctx, "failed to accept X11 connection", slog.Error(err))
+			return
 		}
-	}()
 
-	return display, true
+		// Update session usage time since a new X11 connection was forwarded.
+		x.mu.Lock()
+		session.usedAt = time.Now()
+		x.mu.Unlock()
+		if x11.SingleConnection {
+			x.logger.Debug(ctx, "single connection requested, closing X11 listener")
+			x.closeAndRemoveSession(session)
+		}
+
+		var originAddr string
+		var originPort uint32
+
+		if tcpConn, ok := conn.(*net.TCPConn); ok {
+			if tcpAddr, ok := tcpConn.LocalAddr().(*net.TCPAddr); ok && tcpAddr != nil {
+				originAddr = tcpAddr.IP.String()
+				// #nosec G115 - Safe conversion as TCP port numbers are within uint32 range (0-65535)
+				originPort = uint32(tcpAddr.Port)
+			}
+		}
+		// Fallback values for in-memory or non-TCP connections.
+		if originAddr == "" {
+			originAddr = "127.0.0.1"
+		}
+
+		channel, reqs, err := serverConn.OpenChannel("x11", gossh.Marshal(struct {
+			OriginatorAddress string
+			OriginatorPort    uint32
+		}{
+			OriginatorAddress: originAddr,
+			OriginatorPort:    originPort,
+		}))
+		if err != nil {
+			x.logger.Warn(ctx, "failed to open X11 channel", slog.Error(err))
+			_ = conn.Close()
+			continue
+		}
+		go gossh.DiscardRequests(reqs)
+
+		if !x.trackConn(conn, true) {
+			x.logger.Warn(ctx, "failed to track X11 connection")
+			_ = conn.Close()
+			continue
+		}
+		go func() {
+			defer x.trackConn(conn, false)
+			Bicopy(ctx, conn, channel)
+		}()
+	}
+}
+
+// closeAndRemoveSession closes and removes the session.
+func (x *x11Forwarder) closeAndRemoveSession(x11session *x11Session) {
+	_ = x11session.listener.Close()
+	x.mu.Lock()
+	delete(x.sessions, x11session)
+	x.mu.Unlock()
+}
+
+// createX11Session creates an X11 forwarding session.
+func (x *x11Forwarder) createX11Session(ctx context.Context, sshSession ssh.Session) (*x11Session, error) {
+	var (
+		ln      net.Listener
+		display int
+		err     error
+	)
+	// retry listener creation after evictions. Limit to 10 retries to prevent pathological cases looping forever.
+	const maxRetries = 10
+	for try := range maxRetries {
+		ln, display, err = x.createX11Listener(ctx)
+		if err == nil {
+			break
+		}
+		if try == maxRetries-1 {
+			return nil, xerrors.New("max retries exceeded while creating X11 session")
+		}
+		x.logger.Warn(ctx, "failed to create X11 listener; will evict an X11 forwarding session",
+			slog.F("num_current_sessions", x.numSessions()),
+			slog.Error(err))
+		x.evictLeastRecentlyUsedSession()
+	}
+	x.mu.Lock()
+	defer x.mu.Unlock()
+	if x.closing {
+		closeErr := ln.Close()
+		if closeErr != nil {
+			x.logger.Error(ctx, "error closing X11 listener", slog.Error(closeErr))
+		}
+		return nil, xerrors.New("server is closing")
+	}
+	x11Sess := &x11Session{
+		session:  sshSession,
+		display:  display,
+		listener: ln,
+		usedAt:   time.Now(),
+	}
+	x.sessions[x11Sess] = struct{}{}
+	return x11Sess, nil
+}
+
+func (x *x11Forwarder) numSessions() int {
+	x.mu.Lock()
+	defer x.mu.Unlock()
+	return len(x.sessions)
+}
+
+func (x *x11Forwarder) popLeastRecentlyUsedSession() *x11Session {
+	x.mu.Lock()
+	defer x.mu.Unlock()
+	var lru *x11Session
+	for s := range x.sessions {
+		if lru == nil {
+			lru = s
+			continue
+		}
+		if s.usedAt.Before(lru.usedAt) {
+			lru = s
+			continue
+		}
+	}
+	if lru == nil {
+		x.logger.Debug(context.Background(), "tried to pop from empty set of X11 sessions")
+		return nil
+	}
+	delete(x.sessions, lru)
+	return lru
+}
+
+func (x *x11Forwarder) evictLeastRecentlyUsedSession() {
+	lru := x.popLeastRecentlyUsedSession()
+	if lru == nil {
+		return
+	}
+	err := lru.listener.Close()
+	if err != nil {
+		x.logger.Error(context.Background(), "failed to close evicted X11 session listener", slog.Error(err))
+	}
+	// when we evict, we also want to force the SSH session to be closed as well. This is because we intend to reuse
+	// the X11 TCP listener port for a new X11 forwarding session. If we left the SSH session up, then graphical apps
+	// started in that session could potentially connect to an unintended X11 Server (i.e. the display on a different
+	// computer than the one that started the SSH session). Most likely, this session is a zombie anyway if we've
+	// reached the maximum number of X11 forwarding sessions.
+	err = lru.session.Close()
+	if err != nil {
+		x.logger.Error(context.Background(), "failed to close evicted X11 SSH session", slog.Error(err))
+	}
 }
 
 // createX11Listener creates a listener for X11 forwarding, it will use
 // the next available port starting from X11StartPort and displayOffset.
-func createX11Listener(ctx context.Context, displayOffset int) (ln net.Listener, display int, err error) {
-	var lc net.ListenConfig
+func (x *x11Forwarder) createX11Listener(ctx context.Context) (ln net.Listener, display int, err error) {
 	// Look for an open port to listen on.
-	for port := X11StartPort + displayOffset; port < math.MaxUint16; port++ {
-		ln, err = lc.Listen(ctx, "tcp", fmt.Sprintf("localhost:%d", port))
+	for port := X11StartPort + x.displayOffset; port <= X11MaxPort; port++ {
+		if ctx.Err() != nil {
+			return nil, -1, ctx.Err()
+		}
+
+		ln, err = x.network.Listen("tcp", fmt.Sprintf("localhost:%d", port))
 		if err == nil {
 			display = port - X11StartPort
 			return ln, display, nil
@@ -156,6 +328,49 @@ func createX11Listener(ctx context.Context, displayOffset int) (ln net.Listener,
 	return nil, -1, xerrors.Errorf("failed to find open port for X11 listener: %w", err)
 }
 
+// trackConn registers the connection with the x11Forwarder. If the server is
+// closed, the connection is not registered and should be closed.
+//
+//nolint:revive
+func (x *x11Forwarder) trackConn(c net.Conn, add bool) (ok bool) {
+	x.mu.Lock()
+	defer x.mu.Unlock()
+	if add {
+		if x.closing {
+			// Server or listener closed.
+			return false
+		}
+		x.wg.Add(1)
+		x.connections[c] = struct{}{}
+		return true
+	}
+	x.wg.Done()
+	delete(x.connections, c)
+	return true
+}
+
+func (x *x11Forwarder) Close() error {
+	x.mu.Lock()
+	x.closing = true
+
+	for s := range x.sessions {
+		sErr := s.listener.Close()
+		if sErr != nil {
+			x.logger.Debug(context.Background(), "failed to close X11 listener", slog.Error(sErr))
+		}
+	}
+	for c := range x.connections {
+		cErr := c.Close()
+		if cErr != nil {
+			x.logger.Debug(context.Background(), "failed to close X11 connection", slog.Error(cErr))
+		}
+	}
+
+	x.mu.Unlock()
+	x.wg.Wait()
+	return nil
+}
+
 // addXauthEntry adds an Xauthority entry to the Xauthority file.
 // The Xauthority file is located at ~/.Xauthority.
 func addXauthEntry(ctx context.Context, fs afero.Fs, host string, display string, authProtocol string, authCookie string) error {

agent/agentssh/x11_test.go

@@ -3,9 +3,9 @@ package agentssh_test
 import (
 	"bufio"
 	"bytes"
-	"context"
 	"encoding/hex"
 	"fmt"
+	"io"
 	"net"
 	"os"
 	"path/filepath"
@@ -32,10 +32,19 @@ func TestServer_X11(t *testing.T) {
 		t.Skip("X11 forwarding is only supported on Linux")
 	}
 
-	ctx := context.Background()
+	ctx := testutil.Context(t, testutil.WaitShort)
 	logger := testutil.Logger(t)
-	fs := afero.NewOsFs()
-	s, err := agentssh.NewServer(ctx, logger, prometheus.NewRegistry(), fs, agentexec.DefaultExecer, &agentssh.Config{})
+	fs := afero.NewMemMapFs()
+
+	// Use in-process networking for X11 forwarding.
+	inproc := testutil.NewInProcNet()
+
+	// Create server config with custom X11 listener.
+	cfg := &agentssh.Config{
+		X11Net: inproc,
+	}
+
+	s, err := agentssh.NewServer(ctx, logger, prometheus.NewRegistry(), fs, agentexec.DefaultExecer, cfg)
 	require.NoError(t, err)
 	defer s.Close()
 	err = s.UpdateHostSigner(42)
@@ -93,17 +102,15 @@ func TestServer_X11(t *testing.T) {
 
 	x11Chans := c.HandleChannelOpen("x11")
 	payload := "hello world"
-	require.Eventually(t, func() bool {
-		conn, err := net.Dial("tcp", fmt.Sprintf("localhost:%d", agentssh.X11StartPort+displayNumber))
-		if err == nil {
-			_, err = conn.Write([]byte(payload))
-			assert.NoError(t, err)
-			_ = conn.Close()
-		}
-		return err == nil
-	}, testutil.WaitShort, testutil.IntervalFast)
+	go func() {
+		conn, err := inproc.Dial(ctx, testutil.NewAddr("tcp", fmt.Sprintf("localhost:%d", agentssh.X11StartPort+displayNumber)))
+		assert.NoError(t, err)
+		_, err = conn.Write([]byte(payload))
+		assert.NoError(t, err)
+		_ = conn.Close()
+	}()
 
-	x11 := <-x11Chans
+	x11 := testutil.RequireReceive(ctx, t, x11Chans)
 	ch, reqs, err := x11.Accept()
 	require.NoError(t, err)
 	go gossh.DiscardRequests(reqs)
@@ -121,3 +128,211 @@ func TestServer_X11(t *testing.T) {
 	_, err = fs.Stat(filepath.Join(home, ".Xauthority"))
 	require.NoError(t, err)
 }
+
+func TestServer_X11_EvictionLRU(t *testing.T) {
+	t.Parallel()
+	if runtime.GOOS != "linux" {
+		t.Skip("X11 forwarding is only supported on Linux")
+	}
+
+	ctx := testutil.Context(t, testutil.WaitSuperLong)
+	logger := testutil.Logger(t)
+	fs := afero.NewMemMapFs()
+
+	// Use in-process networking for X11 forwarding.
+	inproc := testutil.NewInProcNet()
+
+	cfg := &agentssh.Config{
+		X11Net: inproc,
+	}
+
+	s, err := agentssh.NewServer(ctx, logger, prometheus.NewRegistry(), fs, agentexec.DefaultExecer, cfg)
+	require.NoError(t, err)
+	defer s.Close()
+	err = s.UpdateHostSigner(42)
+	require.NoError(t, err)
+
+	ln, err := net.Listen("tcp", "127.0.0.1:0")
+	require.NoError(t, err)
+
+	done := testutil.Go(t, func() {
+		err := s.Serve(ln)
+		assert.Error(t, err)
+	})
+
+	c := sshClient(t, ln.Addr().String())
+
+	// block off one port to test x11Forwarder evicts at highest port, not number of listeners.
+	externalListener, err := inproc.Listen("tcp",
+		fmt.Sprintf("localhost:%d", agentssh.X11StartPort+agentssh.X11DefaultDisplayOffset+1))
+	require.NoError(t, err)
+	defer externalListener.Close()
+
+	// Calculate how many simultaneous X11 sessions we can create given the
+	// configured port range.
+
+	startPort := agentssh.X11StartPort + agentssh.X11DefaultDisplayOffset
+	maxSessions := agentssh.X11MaxPort - startPort + 1 - 1 // -1 for the blocked port
+	require.Greater(t, maxSessions, 0, "expected a positive maxSessions value")
+
+	// shellSession holds references to the session and its standard streams so
+	// that the test can keep them open (and optionally interact with them) for
+	// the lifetime of the test. If we don't start the Shell with pipes in place,
+	// the session will be torn down asynchronously during the test.
+	type shellSession struct {
+		sess   *gossh.Session
+		stdin  io.WriteCloser
+		stdout io.Reader
+		stderr io.Reader
+		// scanner is used to read the output of the session, line by line.
+		scanner *bufio.Scanner
+	}
+
+	sessions := make([]shellSession, 0, maxSessions)
+	for i := 0; i < maxSessions; i++ {
+		sess, err := c.NewSession()
+		require.NoError(t, err)
+
+		_, err = sess.SendRequest("x11-req", true, gossh.Marshal(ssh.X11{
+			AuthProtocol: "MIT-MAGIC-COOKIE-1",
+			AuthCookie:   hex.EncodeToString([]byte(fmt.Sprintf("cookie%d", i))),
+			ScreenNumber: uint32(0),
+		}))
+		require.NoError(t, err)
+
+		stdin, err := sess.StdinPipe()
+		require.NoError(t, err)
+		stdout, err := sess.StdoutPipe()
+		require.NoError(t, err)
+		stderr, err := sess.StderrPipe()
+		require.NoError(t, err)
+		require.NoError(t, sess.Shell())
+
+		// The SSH server lazily starts the session. We need to write a command
+		// and read back to ensure the X11 forwarding is started.
+		scanner := bufio.NewScanner(stdout)
+		msg := fmt.Sprintf("ready-%d", i)
+		_, err = stdin.Write([]byte("echo " + msg + "\n"))
+		require.NoError(t, err)
+		// Read until we get the message (first token may be empty due to shell prompt)
+		for scanner.Scan() {
+			line := strings.TrimSpace(scanner.Text())
+			if strings.Contains(line, msg) {
+				break
+			}
+		}
+		require.NoError(t, scanner.Err())
+
+		sessions = append(sessions, shellSession{
+			sess:    sess,
+			stdin:   stdin,
+			stdout:  stdout,
+			stderr:  stderr,
+			scanner: scanner,
+		})
+	}
+
+	// Connect X11 forwarding to the first session. This is used to test that
+	// connecting counts as a use of the display.
+	x11Chans := c.HandleChannelOpen("x11")
+	payload := "hello world"
+	go func() {
+		conn, err := inproc.Dial(ctx, testutil.NewAddr("tcp", fmt.Sprintf("localhost:%d", agentssh.X11StartPort+agentssh.X11DefaultDisplayOffset)))
+		if !assert.NoError(t, err) {
+			return
+		}
+		_, err = conn.Write([]byte(payload))
+		assert.NoError(t, err)
+		_ = conn.Close()
+	}()
+
+	x11 := testutil.RequireReceive(ctx, t, x11Chans)
+	ch, reqs, err := x11.Accept()
+	require.NoError(t, err)
+	go gossh.DiscardRequests(reqs)
+	got := make([]byte, len(payload))
+	_, err = ch.Read(got)
+	require.NoError(t, err)
+	assert.Equal(t, payload, string(got))
+	_ = ch.Close()
+
+	// Create one more session which should evict a session and reuse the display.
+	// The first session was used to connect X11 forwarding, so it should not be evicted.
+	// Therefore, the second session should be evicted and its display reused.
+	extraSess, err := c.NewSession()
+	require.NoError(t, err)
+
+	_, err = extraSess.SendRequest("x11-req", true, gossh.Marshal(ssh.X11{
+		AuthProtocol: "MIT-MAGIC-COOKIE-1",
+		AuthCookie:   hex.EncodeToString([]byte("extra")),
+		ScreenNumber: uint32(0),
+	}))
+	require.NoError(t, err)
+
+	// Ask the remote side for the DISPLAY value so we can extract the display
+	// number that was assigned to this session.
+	out, err := extraSess.Output("echo DISPLAY=$DISPLAY")
+	require.NoError(t, err)
+
+	// Example output line: "DISPLAY=localhost:10.0".
+	var newDisplayNumber int
+	{
+		sc := bufio.NewScanner(bytes.NewReader(out))
+		for sc.Scan() {
+			line := strings.TrimSpace(sc.Text())
+			if strings.HasPrefix(line, "DISPLAY=") {
+				parts := strings.SplitN(line, ":", 2)
+				require.Len(t, parts, 2)
+				displayPart := parts[1]
+				if strings.Contains(displayPart, ".") {
+					displayPart = strings.SplitN(displayPart, ".", 2)[0]
+				}
+				var convErr error
+				newDisplayNumber, convErr = strconv.Atoi(displayPart)
+				require.NoError(t, convErr)
+				break
+			}
+		}
+		require.NoError(t, sc.Err())
+	}
+
+	// The display number reused should correspond to the SECOND session (display offset 12)
+	expectedDisplay := agentssh.X11DefaultDisplayOffset + 2 // +1 was blocked port
+	assert.Equal(t, expectedDisplay, newDisplayNumber, "second session should have been evicted and its display reused")
+
+	// First session should still be alive: send cmd and read output.
+	msgFirst := "still-alive"
+	_, err = sessions[0].stdin.Write([]byte("echo " + msgFirst + "\n"))
+	require.NoError(t, err)
+	for sessions[0].scanner.Scan() {
+		line := strings.TrimSpace(sessions[0].scanner.Text())
+		if strings.Contains(line, msgFirst) {
+			break
+		}
+	}
+	require.NoError(t, sessions[0].scanner.Err())
+
+	// Second session should now be closed.
+	_, err = sessions[1].stdin.Write([]byte("echo dead\n"))
+	require.ErrorIs(t, err, io.EOF)
+	err = sessions[1].sess.Wait()
+	require.Error(t, err)
+
+	// Cleanup.
+	for i, sh := range sessions {
+		if i == 1 {
+			// already closed
+			continue
+		}
+		err = sh.stdin.Close()
+		require.NoError(t, err)
+		err = sh.sess.Wait()
+		require.NoError(t, err)
+	}
+	err = extraSess.Close()
+	require.ErrorIs(t, err, io.EOF)
+
+	err = s.Close()
+	require.NoError(t, err)
+	_ = testutil.TryReceive(ctx, t, done)
+}

agent/agenttest/agent.go

@@ -1,7 +1,6 @@
 package agenttest
 
 import (
-	"context"
 	"net/url"
 	"testing"
 
@@ -31,18 +30,11 @@ func New(t testing.TB, coderURL *url.URL, agentToken string, opts ...func(*agent
 	}
 
 	if o.Client == nil {
-		agentClient := agentsdk.New(coderURL)
-		agentClient.SetSessionToken(agentToken)
+		agentClient := agentsdk.New(coderURL, agentsdk.WithFixedToken(agentToken))
 		agentClient.SDK.SetLogger(log)
 		o.Client = agentClient
 	}
 
-	if o.ExchangeToken == nil {
-		o.ExchangeToken = func(_ context.Context) (string, error) {
-			return agentToken, nil
-		}
-	}
-
 	if o.LogDir == "" {
 		o.LogDir = t.TempDir()
 	}

agent/agenttest/client.go

@@ -3,6 +3,7 @@ package agenttest
 import (
 	"context"
 	"io"
+	"net/http"
 	"slices"
 	"sync"
 	"sync/atomic"
@@ -28,6 +29,7 @@ import (
 	"github.com/coder/coder/v2/tailnet"
 	"github.com/coder/coder/v2/tailnet/proto"
 	"github.com/coder/coder/v2/testutil"
+	"github.com/coder/websocket"
 )
 
 const statsInterval = 500 * time.Millisecond
@@ -86,10 +88,34 @@ type Client struct {
 	fakeAgentAPI       *FakeAgentAPI
 	LastWorkspaceAgent func()
 
-	mu             sync.Mutex // Protects following.
-	logs           []agentsdk.Log
-	derpMapUpdates chan *tailcfg.DERPMap
-	derpMapOnce    sync.Once
+	mu                sync.Mutex // Protects following.
+	logs              []agentsdk.Log
+	derpMapUpdates    chan *tailcfg.DERPMap
+	derpMapOnce       sync.Once
+	refreshTokenCalls int
+}
+
+func (*Client) AsRequestOption() codersdk.RequestOption {
+	return func(_ *http.Request) {}
+}
+
+func (*Client) SetDialOption(*websocket.DialOptions) {}
+
+func (*Client) GetSessionToken() string {
+	return "agenttest-token"
+}
+
+func (c *Client) RefreshToken(context.Context) error {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	c.refreshTokenCalls++
+	return nil
+}
+
+func (c *Client) GetNumRefreshTokenCalls() int {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	return c.refreshTokenCalls
 }
 
 func (*Client) RewriteDERPMap(*tailcfg.DERPMap) {}

agent/api.go

@@ -2,46 +2,44 @@ package agent
 
 import (
 	"net/http"
-	"sync"
-	"time"
 
 	"github.com/go-chi/chi/v5"
+	"github.com/google/uuid"
 
 	"github.com/coder/coder/v2/coderd/httpapi"
+	"github.com/coder/coder/v2/coderd/httpmw/loggermw"
+	"github.com/coder/coder/v2/coderd/tracing"
 	"github.com/coder/coder/v2/codersdk"
+	"github.com/coder/coder/v2/codersdk/workspacesdk"
+	"github.com/coder/coder/v2/httpmw"
 )
 
 func (a *agent) apiHandler() http.Handler {
 	r := chi.NewRouter()
+	r.Use(
+		httpmw.Recover(a.logger),
+		tracing.StatusWriterMiddleware,
+		loggermw.Logger(a.logger),
+	)
 	r.Get("/", func(rw http.ResponseWriter, r *http.Request) {
 		httpapi.Write(r.Context(), rw, http.StatusOK, codersdk.Response{
 			Message: "Hello from the agent!",
 		})
 	})
 
-	// Make a copy to ensure the map is not modified after the handler is
-	// created.
-	cpy := make(map[int]string)
-	for k, b := range a.ignorePorts {
-		cpy[k] = b
-	}
-
-	cacheDuration := 1 * time.Second
-	if a.portCacheDuration > 0 {
-		cacheDuration = a.portCacheDuration
-	}
-
-	lp := &listeningPortsHandler{
-		ignorePorts:   cpy,
-		cacheDuration: cacheDuration,
-	}
-
-	if a.containerAPI != nil {
+	if a.devcontainers {
 		r.Mount("/api/v0/containers", a.containerAPI.Routes())
+	} else if manifest := a.manifest.Load(); manifest != nil && manifest.ParentID != uuid.Nil {
+		r.HandleFunc("/api/v0/containers", func(w http.ResponseWriter, r *http.Request) {
+			httpapi.Write(r.Context(), w, http.StatusForbidden, codersdk.Response{
+				Message: "Dev Container feature not supported.",
+				Detail:  "Dev Container integration inside other Dev Containers is explicitly not supported.",
+			})
+		})
 	} else {
 		r.HandleFunc("/api/v0/containers", func(w http.ResponseWriter, r *http.Request) {
 			httpapi.Write(r.Context(), w, http.StatusForbidden, codersdk.Response{
-				Message: "The agent dev containers feature is experimental and not enabled by default.",
+				Message: "Dev Container feature not enabled.",
 				Detail:  "To enable this feature, set CODER_AGENT_DEVCONTAINERS_ENABLE=true in your template.",
 			})
 		})
@@ -49,9 +47,12 @@ func (a *agent) apiHandler() http.Handler {
 
 	promHandler := PrometheusMetricsHandler(a.prometheusRegistry, a.logger)
 
-	r.Get("/api/v0/listening-ports", lp.handler)
+	r.Get("/api/v0/listening-ports", a.listeningPortsHandler.handler)
 	r.Get("/api/v0/netcheck", a.HandleNetcheck)
 	r.Post("/api/v0/list-directory", a.HandleLS)
+	r.Get("/api/v0/read-file", a.HandleReadFile)
+	r.Post("/api/v0/write-file", a.HandleWriteFile)
+	r.Post("/api/v0/edit-files", a.HandleEditFiles)
 	r.Get("/debug/logs", a.HandleHTTPDebugLogs)
 	r.Get("/debug/magicsock", a.HandleHTTPDebugMagicsock)
 	r.Get("/debug/magicsock/debug-logging/{state}", a.HandleHTTPMagicsockDebugLoggingState)
@@ -61,22 +62,21 @@ func (a *agent) apiHandler() http.Handler {
 	return r
 }
 
-type listeningPortsHandler struct {
-	ignorePorts   map[int]string
-	cacheDuration time.Duration
+type ListeningPortsGetter interface {
+	GetListeningPorts() ([]codersdk.WorkspaceAgentListeningPort, error)
+}
 
-	//nolint: unused  // used on some but not all platforms
-	mut sync.Mutex
-	//nolint: unused  // used on some but not all platforms
-	ports []codersdk.WorkspaceAgentListeningPort
-	//nolint: unused  // used on some but not all platforms
-	mtime time.Time
+type listeningPortsHandler struct {
+	// In production code, this is set to an osListeningPortsGetter, but it can be overridden for
+	// testing.
+	getter      ListeningPortsGetter
+	ignorePorts map[int]string
 }
 
 // handler returns a list of listening ports. This is tested by coderd's
 // TestWorkspaceAgentListeningPorts test.
 func (lp *listeningPortsHandler) handler(rw http.ResponseWriter, r *http.Request) {
-	ports, err := lp.getListeningPorts()
+	ports, err := lp.getter.GetListeningPorts()
 	if err != nil {
 		httpapi.Write(r.Context(), rw, http.StatusInternalServerError, codersdk.Response{
 			Message: "Could not scan for listening ports.",
@@ -85,7 +85,20 @@ func (lp *listeningPortsHandler) handler(rw http.ResponseWriter, r *http.Request
 		return
 	}
 
+	filteredPorts := make([]codersdk.WorkspaceAgentListeningPort, 0, len(ports))
+	for _, port := range ports {
+		if port.Port < workspacesdk.AgentMinimumListeningPort {
+			continue
+		}
+
+		// Ignore ports that we've been told to ignore.
+		if _, ok := lp.ignorePorts[int(port.Port)]; ok {
+			continue
+		}
+		filteredPorts = append(filteredPorts, port)
+	}
+
 	httpapi.Write(r.Context(), rw, http.StatusOK, codersdk.WorkspaceAgentListeningPortsResponse{
-		Ports: ports,
+		Ports: filteredPorts,
 	})
 }

agent/apphealth.go

@@ -63,6 +63,7 @@ func NewAppHealthReporterWithClock(
 		// run a ticker for each app health check.
 		var mu sync.RWMutex
 		failures := make(map[uuid.UUID]int, 0)
+		client := &http.Client{}
 		for _, nextApp := range apps {
 			if !shouldStartTicker(nextApp) {
 				continue
@@ -91,7 +92,7 @@ func NewAppHealthReporterWithClock(
 						if err != nil {
 							return err
 						}
-						res, err := http.DefaultClient.Do(req)
+						res, err := client.Do(req)
 						if err != nil {
 							return err
 						}

agent/files.go

@@ -0,0 +1,275 @@
+package agent
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"mime"
+	"net/http"
+	"os"
+	"path/filepath"
+	"strconv"
+	"syscall"
+
+	"github.com/icholy/replace"
+	"github.com/spf13/afero"
+	"golang.org/x/text/transform"
+	"golang.org/x/xerrors"
+
+	"cdr.dev/slog"
+	"github.com/coder/coder/v2/coderd/httpapi"
+	"github.com/coder/coder/v2/codersdk"
+	"github.com/coder/coder/v2/codersdk/workspacesdk"
+)
+
+type HTTPResponseCode = int
+
+func (a *agent) HandleReadFile(rw http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	query := r.URL.Query()
+	parser := httpapi.NewQueryParamParser().RequiredNotEmpty("path")
+	path := parser.String(query, "", "path")
+	offset := parser.PositiveInt64(query, 0, "offset")
+	limit := parser.PositiveInt64(query, 0, "limit")
+	parser.ErrorExcessParams(query)
+	if len(parser.Errors) > 0 {
+		httpapi.Write(ctx, rw, http.StatusBadRequest, codersdk.Response{
+			Message:     "Query parameters have invalid values.",
+			Validations: parser.Errors,
+		})
+		return
+	}
+
+	status, err := a.streamFile(ctx, rw, path, offset, limit)
+	if err != nil {
+		httpapi.Write(ctx, rw, status, codersdk.Response{
+			Message: err.Error(),
+		})
+		return
+	}
+}
+
+func (a *agent) streamFile(ctx context.Context, rw http.ResponseWriter, path string, offset, limit int64) (HTTPResponseCode, error) {
+	if !filepath.IsAbs(path) {
+		return http.StatusBadRequest, xerrors.Errorf("file path must be absolute: %q", path)
+	}
+
+	f, err := a.filesystem.Open(path)
+	if err != nil {
+		status := http.StatusInternalServerError
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+			status = http.StatusNotFound
+		case errors.Is(err, os.ErrPermission):
+			status = http.StatusForbidden
+		}
+		return status, err
+	}
+	defer f.Close()
+
+	stat, err := f.Stat()
+	if err != nil {
+		return http.StatusInternalServerError, err
+	}
+
+	if stat.IsDir() {
+		return http.StatusBadRequest, xerrors.Errorf("open %s: not a file", path)
+	}
+
+	size := stat.Size()
+	if limit == 0 {
+		limit = size
+	}
+	bytesRemaining := max(size-offset, 0)
+	bytesToRead := min(bytesRemaining, limit)
+
+	// Relying on just the file name for the mime type for now.
+	mimeType := mime.TypeByExtension(filepath.Ext(path))
+	if mimeType == "" {
+		mimeType = "application/octet-stream"
+	}
+	rw.Header().Set("Content-Type", mimeType)
+	rw.Header().Set("Content-Length", strconv.FormatInt(bytesToRead, 10))
+	rw.WriteHeader(http.StatusOK)
+
+	reader := io.NewSectionReader(f, offset, bytesToRead)
+	_, err = io.Copy(rw, reader)
+	if err != nil && !errors.Is(err, io.EOF) && ctx.Err() == nil {
+		a.logger.Error(ctx, "workspace agent read file", slog.Error(err))
+	}
+
+	return 0, nil
+}
+
+func (a *agent) HandleWriteFile(rw http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	query := r.URL.Query()
+	parser := httpapi.NewQueryParamParser().RequiredNotEmpty("path")
+	path := parser.String(query, "", "path")
+	parser.ErrorExcessParams(query)
+	if len(parser.Errors) > 0 {
+		httpapi.Write(ctx, rw, http.StatusBadRequest, codersdk.Response{
+			Message:     "Query parameters have invalid values.",
+			Validations: parser.Errors,
+		})
+		return
+	}
+
+	status, err := a.writeFile(ctx, r, path)
+	if err != nil {
+		httpapi.Write(ctx, rw, status, codersdk.Response{
+			Message: err.Error(),
+		})
+		return
+	}
+
+	httpapi.Write(ctx, rw, http.StatusOK, codersdk.Response{
+		Message: fmt.Sprintf("Successfully wrote to %q", path),
+	})
+}
+
+func (a *agent) writeFile(ctx context.Context, r *http.Request, path string) (HTTPResponseCode, error) {
+	if !filepath.IsAbs(path) {
+		return http.StatusBadRequest, xerrors.Errorf("file path must be absolute: %q", path)
+	}
+
+	dir := filepath.Dir(path)
+	err := a.filesystem.MkdirAll(dir, 0o755)
+	if err != nil {
+		status := http.StatusInternalServerError
+		switch {
+		case errors.Is(err, os.ErrPermission):
+			status = http.StatusForbidden
+		case errors.Is(err, syscall.ENOTDIR):
+			status = http.StatusBadRequest
+		}
+		return status, err
+	}
+
+	f, err := a.filesystem.Create(path)
+	if err != nil {
+		status := http.StatusInternalServerError
+		switch {
+		case errors.Is(err, os.ErrPermission):
+			status = http.StatusForbidden
+		case errors.Is(err, syscall.EISDIR):
+			status = http.StatusBadRequest
+		}
+		return status, err
+	}
+	defer f.Close()
+
+	_, err = io.Copy(f, r.Body)
+	if err != nil && !errors.Is(err, io.EOF) && ctx.Err() == nil {
+		a.logger.Error(ctx, "workspace agent write file", slog.Error(err))
+	}
+
+	return 0, nil
+}
+
+func (a *agent) HandleEditFiles(rw http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	var req workspacesdk.FileEditRequest
+	if !httpapi.Read(ctx, rw, r, &req) {
+		return
+	}
+
+	if len(req.Files) == 0 {
+		httpapi.Write(ctx, rw, http.StatusBadRequest, codersdk.Response{
+			Message: "must specify at least one file",
+		})
+		return
+	}
+
+	var combinedErr error
+	status := http.StatusOK
+	for _, edit := range req.Files {
+		s, err := a.editFile(r.Context(), edit.Path, edit.Edits)
+		// Keep the highest response status, so 500 will be preferred over 400, etc.
+		if s > status {
+			status = s
+		}
+		if err != nil {
+			combinedErr = errors.Join(combinedErr, err)
+		}
+	}
+
+	if combinedErr != nil {
+		httpapi.Write(ctx, rw, status, codersdk.Response{
+			Message: combinedErr.Error(),
+		})
+		return
+	}
+
+	httpapi.Write(ctx, rw, http.StatusOK, codersdk.Response{
+		Message: "Successfully edited file(s)",
+	})
+}
+
+func (a *agent) editFile(ctx context.Context, path string, edits []workspacesdk.FileEdit) (int, error) {
+	if path == "" {
+		return http.StatusBadRequest, xerrors.New("\"path\" is required")
+	}
+
+	if !filepath.IsAbs(path) {
+		return http.StatusBadRequest, xerrors.Errorf("file path must be absolute: %q", path)
+	}
+
+	if len(edits) == 0 {
+		return http.StatusBadRequest, xerrors.New("must specify at least one edit")
+	}
+
+	f, err := a.filesystem.Open(path)
+	if err != nil {
+		status := http.StatusInternalServerError
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+			status = http.StatusNotFound
+		case errors.Is(err, os.ErrPermission):
+			status = http.StatusForbidden
+		}
+		return status, err
+	}
+	defer f.Close()
+
+	stat, err := f.Stat()
+	if err != nil {
+		return http.StatusInternalServerError, err
+	}
+
+	if stat.IsDir() {
+		return http.StatusBadRequest, xerrors.Errorf("open %s: not a file", path)
+	}
+
+	transforms := make([]transform.Transformer, len(edits))
+	for i, edit := range edits {
+		transforms[i] = replace.String(edit.Search, edit.Replace)
+	}
+
+	// Create an adjacent file to ensure it will be on the same device and can be
+	// moved atomically.
+	tmpfile, err := afero.TempFile(a.filesystem, filepath.Dir(path), filepath.Base(path))
+	if err != nil {
+		return http.StatusInternalServerError, err
+	}
+	defer tmpfile.Close()
+
+	_, err = io.Copy(tmpfile, replace.Chain(f, transforms...))
+	if err != nil {
+		if rerr := a.filesystem.Remove(tmpfile.Name()); rerr != nil {
+			a.logger.Warn(ctx, "unable to clean up temp file", slog.Error(rerr))
+		}
+		return http.StatusInternalServerError, xerrors.Errorf("edit %s: %w", path, err)
+	}
+
+	err = a.filesystem.Rename(tmpfile.Name(), path)
+	if err != nil {
+		return http.StatusInternalServerError, err
+	}
+
+	return 0, nil
+}

agent/files_test.go

@@ -0,0 +1,722 @@
+package agent_test
+
+import (
+	"bytes"
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+	"os"
+	"path/filepath"
+	"runtime"
+	"syscall"
+	"testing"
+
+	"github.com/spf13/afero"
+	"github.com/stretchr/testify/require"
+	"golang.org/x/xerrors"
+
+	"github.com/coder/coder/v2/agent"
+	"github.com/coder/coder/v2/agent/agenttest"
+	"github.com/coder/coder/v2/coderd/coderdtest"
+	"github.com/coder/coder/v2/codersdk/agentsdk"
+	"github.com/coder/coder/v2/codersdk/workspacesdk"
+	"github.com/coder/coder/v2/testutil"
+)
+
+type testFs struct {
+	afero.Fs
+	// intercept can return an error for testing when a call fails.
+	intercept func(call, file string) error
+}
+
+func newTestFs(base afero.Fs, intercept func(call, file string) error) *testFs {
+	return &testFs{
+		Fs:        base,
+		intercept: intercept,
+	}
+}
+
+func (fs *testFs) Open(name string) (afero.File, error) {
+	if err := fs.intercept("open", name); err != nil {
+		return nil, err
+	}
+	return fs.Fs.Open(name)
+}
+
+func (fs *testFs) Create(name string) (afero.File, error) {
+	if err := fs.intercept("create", name); err != nil {
+		return nil, err
+	}
+	// Unlike os, afero lets you create files where directories already exist and
+	// lets you nest them underneath files, somehow.
+	stat, err := fs.Fs.Stat(name)
+	if err == nil && stat.IsDir() {
+		return nil, &os.PathError{
+			Op:   "open",
+			Path: name,
+			Err:  syscall.EISDIR,
+		}
+	}
+	stat, err = fs.Fs.Stat(filepath.Dir(name))
+	if err == nil && !stat.IsDir() {
+		return nil, &os.PathError{
+			Op:   "open",
+			Path: name,
+			Err:  syscall.ENOTDIR,
+		}
+	}
+	return fs.Fs.Create(name)
+}
+
+func (fs *testFs) MkdirAll(name string, mode os.FileMode) error {
+	if err := fs.intercept("mkdirall", name); err != nil {
+		return err
+	}
+	// Unlike os, afero lets you create directories where files already exist and
+	// lets you nest them underneath files somehow.
+	stat, err := fs.Fs.Stat(filepath.Dir(name))
+	if err == nil && !stat.IsDir() {
+		return &os.PathError{
+			Op:   "mkdir",
+			Path: name,
+			Err:  syscall.ENOTDIR,
+		}
+	}
+	stat, err = fs.Fs.Stat(name)
+	if err == nil && !stat.IsDir() {
+		return &os.PathError{
+			Op:   "mkdir",
+			Path: name,
+			Err:  syscall.ENOTDIR,
+		}
+	}
+	return fs.Fs.MkdirAll(name, mode)
+}
+
+func (fs *testFs) Rename(oldName, newName string) error {
+	if err := fs.intercept("rename", newName); err != nil {
+		return err
+	}
+	return fs.Fs.Rename(oldName, newName)
+}
+
+func TestReadFile(t *testing.T) {
+	t.Parallel()
+
+	tmpdir := os.TempDir()
+	noPermsFilePath := filepath.Join(tmpdir, "no-perms")
+	//nolint:dogsled
+	conn, _, _, fs, _ := setupAgent(t, agentsdk.Manifest{}, 0, func(_ *agenttest.Client, opts *agent.Options) {
+		opts.Filesystem = newTestFs(opts.Filesystem, func(call, file string) error {
+			if file == noPermsFilePath {
+				return os.ErrPermission
+			}
+			return nil
+		})
+	})
+
+	dirPath := filepath.Join(tmpdir, "a-directory")
+	err := fs.MkdirAll(dirPath, 0o755)
+	require.NoError(t, err)
+
+	filePath := filepath.Join(tmpdir, "file")
+	err = afero.WriteFile(fs, filePath, []byte("content"), 0o644)
+	require.NoError(t, err)
+
+	imagePath := filepath.Join(tmpdir, "file.png")
+	err = afero.WriteFile(fs, imagePath, []byte("not really an image"), 0o644)
+	require.NoError(t, err)
+
+	tests := []struct {
+		name     string
+		path     string
+		limit    int64
+		offset   int64
+		bytes    []byte
+		mimeType string
+		errCode  int
+		error    string
+	}{
+		{
+			name:    "NoPath",
+			path:    "",
+			errCode: http.StatusBadRequest,
+			error:   "\"path\" is required",
+		},
+		{
+			name:    "RelativePathDotSlash",
+			path:    "./relative",
+			errCode: http.StatusBadRequest,
+			error:   "file path must be absolute",
+		},
+		{
+			name:    "RelativePath",
+			path:    "also-relative",
+			errCode: http.StatusBadRequest,
+			error:   "file path must be absolute",
+		},
+		{
+			name:    "NegativeLimit",
+			path:    filePath,
+			limit:   -10,
+			errCode: http.StatusBadRequest,
+			error:   "value is negative",
+		},
+		{
+			name:    "NegativeOffset",
+			path:    filePath,
+			offset:  -10,
+			errCode: http.StatusBadRequest,
+			error:   "value is negative",
+		},
+		{
+			name:    "NonExistent",
+			path:    filepath.Join(tmpdir, "does-not-exist"),
+			errCode: http.StatusNotFound,
+			error:   "file does not exist",
+		},
+		{
+			name:    "IsDir",
+			path:    dirPath,
+			errCode: http.StatusBadRequest,
+			error:   "not a file",
+		},
+		{
+			name:    "NoPermissions",
+			path:    noPermsFilePath,
+			errCode: http.StatusForbidden,
+			error:   "permission denied",
+		},
+		{
+			name:     "Defaults",
+			path:     filePath,
+			bytes:    []byte("content"),
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "Limit1",
+			path:     filePath,
+			limit:    1,
+			bytes:    []byte("c"),
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "Offset1",
+			path:     filePath,
+			offset:   1,
+			bytes:    []byte("ontent"),
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "Limit1Offset2",
+			path:     filePath,
+			limit:    1,
+			offset:   2,
+			bytes:    []byte("n"),
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "Limit7Offset0",
+			path:     filePath,
+			limit:    7,
+			offset:   0,
+			bytes:    []byte("content"),
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "Limit100",
+			path:     filePath,
+			limit:    100,
+			bytes:    []byte("content"),
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "Offset7",
+			path:     filePath,
+			offset:   7,
+			bytes:    []byte{},
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "Offset100",
+			path:     filePath,
+			offset:   100,
+			bytes:    []byte{},
+			mimeType: "application/octet-stream",
+		},
+		{
+			name:     "MimeTypePng",
+			path:     imagePath,
+			bytes:    []byte("not really an image"),
+			mimeType: "image/png",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
+			defer cancel()
+
+			reader, mimeType, err := conn.ReadFile(ctx, tt.path, tt.offset, tt.limit)
+			if tt.errCode != 0 {
+				require.Error(t, err)
+				cerr := coderdtest.SDKError(t, err)
+				require.Contains(t, cerr.Error(), tt.error)
+				require.Equal(t, tt.errCode, cerr.StatusCode())
+			} else {
+				require.NoError(t, err)
+				defer reader.Close()
+				bytes, err := io.ReadAll(reader)
+				require.NoError(t, err)
+				require.Equal(t, tt.bytes, bytes)
+				require.Equal(t, tt.mimeType, mimeType)
+			}
+		})
+	}
+}
+
+func TestWriteFile(t *testing.T) {
+	t.Parallel()
+
+	tmpdir := os.TempDir()
+	noPermsFilePath := filepath.Join(tmpdir, "no-perms-file")
+	noPermsDirPath := filepath.Join(tmpdir, "no-perms-dir")
+	//nolint:dogsled
+	conn, _, _, fs, _ := setupAgent(t, agentsdk.Manifest{}, 0, func(_ *agenttest.Client, opts *agent.Options) {
+		opts.Filesystem = newTestFs(opts.Filesystem, func(call, file string) error {
+			if file == noPermsFilePath || file == noPermsDirPath {
+				return os.ErrPermission
+			}
+			return nil
+		})
+	})
+
+	dirPath := filepath.Join(tmpdir, "directory")
+	err := fs.MkdirAll(dirPath, 0o755)
+	require.NoError(t, err)
+
+	filePath := filepath.Join(tmpdir, "file")
+	err = afero.WriteFile(fs, filePath, []byte("content"), 0o644)
+	require.NoError(t, err)
+
+	notDirErr := "not a directory"
+	if runtime.GOOS == "windows" {
+		notDirErr = "cannot find the path"
+	}
+
+	tests := []struct {
+		name    string
+		path    string
+		bytes   []byte
+		errCode int
+		error   string
+	}{
+		{
+			name:    "NoPath",
+			path:    "",
+			errCode: http.StatusBadRequest,
+			error:   "\"path\" is required",
+		},
+		{
+			name:    "RelativePathDotSlash",
+			path:    "./relative",
+			errCode: http.StatusBadRequest,
+			error:   "file path must be absolute",
+		},
+		{
+			name:    "RelativePath",
+			path:    "also-relative",
+			errCode: http.StatusBadRequest,
+			error:   "file path must be absolute",
+		},
+		{
+			name:  "NonExistent",
+			path:  filepath.Join(tmpdir, "/nested/does-not-exist"),
+			bytes: []byte("now it does exist"),
+		},
+		{
+			name:    "IsDir",
+			path:    dirPath,
+			errCode: http.StatusBadRequest,
+			error:   "is a directory",
+		},
+		{
+			name:    "IsNotDir",
+			path:    filepath.Join(filePath, "file2"),
+			errCode: http.StatusBadRequest,
+			error:   notDirErr,
+		},
+		{
+			name:    "NoPermissionsFile",
+			path:    noPermsFilePath,
+			errCode: http.StatusForbidden,
+			error:   "permission denied",
+		},
+		{
+			name:    "NoPermissionsDir",
+			path:    filepath.Join(noPermsDirPath, "within-no-perm-dir"),
+			errCode: http.StatusForbidden,
+			error:   "permission denied",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
+			defer cancel()
+
+			reader := bytes.NewReader(tt.bytes)
+			err := conn.WriteFile(ctx, tt.path, reader)
+			if tt.errCode != 0 {
+				require.Error(t, err)
+				cerr := coderdtest.SDKError(t, err)
+				require.Contains(t, cerr.Error(), tt.error)
+				require.Equal(t, tt.errCode, cerr.StatusCode())
+			} else {
+				require.NoError(t, err)
+				b, err := afero.ReadFile(fs, tt.path)
+				require.NoError(t, err)
+				require.Equal(t, tt.bytes, b)
+			}
+		})
+	}
+}
+
+func TestEditFiles(t *testing.T) {
+	t.Parallel()
+
+	tmpdir := os.TempDir()
+	noPermsFilePath := filepath.Join(tmpdir, "no-perms-file")
+	failRenameFilePath := filepath.Join(tmpdir, "fail-rename")
+	//nolint:dogsled
+	conn, _, _, fs, _ := setupAgent(t, agentsdk.Manifest{}, 0, func(_ *agenttest.Client, opts *agent.Options) {
+		opts.Filesystem = newTestFs(opts.Filesystem, func(call, file string) error {
+			if file == noPermsFilePath {
+				return &os.PathError{
+					Op:   call,
+					Path: file,
+					Err:  os.ErrPermission,
+				}
+			} else if file == failRenameFilePath && call == "rename" {
+				return xerrors.New("rename failed")
+			}
+			return nil
+		})
+	})
+
+	dirPath := filepath.Join(tmpdir, "directory")
+	err := fs.MkdirAll(dirPath, 0o755)
+	require.NoError(t, err)
+
+	tests := []struct {
+		name     string
+		contents map[string]string
+		edits    []workspacesdk.FileEdits
+		expected map[string]string
+		errCode  int
+		errors   []string
+	}{
+		{
+			name:    "NoFiles",
+			errCode: http.StatusBadRequest,
+			errors:  []string{"must specify at least one file"},
+		},
+		{
+			name:    "NoPath",
+			errCode: http.StatusBadRequest,
+			edits: []workspacesdk.FileEdits{
+				{
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			errors: []string{"\"path\" is required"},
+		},
+		{
+			name: "RelativePathDotSlash",
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: "./relative",
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			errCode: http.StatusBadRequest,
+			errors:  []string{"file path must be absolute"},
+		},
+		{
+			name: "RelativePath",
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: "also-relative",
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			errCode: http.StatusBadRequest,
+			errors:  []string{"file path must be absolute"},
+		},
+		{
+			name: "NoEdits",
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: filepath.Join(tmpdir, "no-edits"),
+				},
+			},
+			errCode: http.StatusBadRequest,
+			errors:  []string{"must specify at least one edit"},
+		},
+		{
+			name: "NonExistent",
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: filepath.Join(tmpdir, "does-not-exist"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			errCode: http.StatusNotFound,
+			errors:  []string{"file does not exist"},
+		},
+		{
+			name: "IsDir",
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: dirPath,
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			errCode: http.StatusBadRequest,
+			errors:  []string{"not a file"},
+		},
+		{
+			name: "NoPermissions",
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: noPermsFilePath,
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			errCode: http.StatusForbidden,
+			errors:  []string{"permission denied"},
+		},
+		{
+			name:     "FailRename",
+			contents: map[string]string{failRenameFilePath: "foo bar"},
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: failRenameFilePath,
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			errCode: http.StatusInternalServerError,
+			errors:  []string{"rename failed"},
+		},
+		{
+			name:     "Edit1",
+			contents: map[string]string{filepath.Join(tmpdir, "edit1"): "foo bar"},
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: filepath.Join(tmpdir, "edit1"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+					},
+				},
+			},
+			expected: map[string]string{filepath.Join(tmpdir, "edit1"): "bar bar"},
+		},
+		{
+			name:     "EditEdit", // Edits affect previous edits.
+			contents: map[string]string{filepath.Join(tmpdir, "edit-edit"): "foo bar"},
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: filepath.Join(tmpdir, "edit-edit"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "foo",
+							Replace: "bar",
+						},
+						{
+							Search:  "bar",
+							Replace: "qux",
+						},
+					},
+				},
+			},
+			expected: map[string]string{filepath.Join(tmpdir, "edit-edit"): "qux qux"},
+		},
+		{
+			name:     "Multiline",
+			contents: map[string]string{filepath.Join(tmpdir, "multiline"): "foo\nbar\nbaz\nqux"},
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: filepath.Join(tmpdir, "multiline"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "bar\nbaz",
+							Replace: "frob",
+						},
+					},
+				},
+			},
+			expected: map[string]string{filepath.Join(tmpdir, "multiline"): "foo\nfrob\nqux"},
+		},
+		{
+			name: "Multifile",
+			contents: map[string]string{
+				filepath.Join(tmpdir, "file1"): "file 1",
+				filepath.Join(tmpdir, "file2"): "file 2",
+				filepath.Join(tmpdir, "file3"): "file 3",
+			},
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: filepath.Join(tmpdir, "file1"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "file",
+							Replace: "edited1",
+						},
+					},
+				},
+				{
+					Path: filepath.Join(tmpdir, "file2"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "file",
+							Replace: "edited2",
+						},
+					},
+				},
+				{
+					Path: filepath.Join(tmpdir, "file3"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "file",
+							Replace: "edited3",
+						},
+					},
+				},
+			},
+			expected: map[string]string{
+				filepath.Join(tmpdir, "file1"): "edited1 1",
+				filepath.Join(tmpdir, "file2"): "edited2 2",
+				filepath.Join(tmpdir, "file3"): "edited3 3",
+			},
+		},
+		{
+			name: "MultiError",
+			contents: map[string]string{
+				filepath.Join(tmpdir, "file8"): "file 8",
+			},
+			edits: []workspacesdk.FileEdits{
+				{
+					Path: noPermsFilePath,
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "file",
+							Replace: "edited7",
+						},
+					},
+				},
+				{
+					Path: filepath.Join(tmpdir, "file8"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "file",
+							Replace: "edited8",
+						},
+					},
+				},
+				{
+					Path: filepath.Join(tmpdir, "file9"),
+					Edits: []workspacesdk.FileEdit{
+						{
+							Search:  "file",
+							Replace: "edited9",
+						},
+					},
+				},
+			},
+			expected: map[string]string{
+				filepath.Join(tmpdir, "file8"): "edited8 8",
+			},
+			// Higher status codes will override lower ones, so in this case the 404
+			// takes priority over the 403.
+			errCode: http.StatusNotFound,
+			errors: []string{
+				fmt.Sprintf("%s: permission denied", noPermsFilePath),
+				"file9: file does not exist",
+			},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
+			defer cancel()
+
+			for path, content := range tt.contents {
+				err := afero.WriteFile(fs, path, []byte(content), 0o644)
+				require.NoError(t, err)
+			}
+
+			err := conn.EditFiles(ctx, workspacesdk.FileEditRequest{Files: tt.edits})
+			if tt.errCode != 0 {
+				require.Error(t, err)
+				cerr := coderdtest.SDKError(t, err)
+				for _, error := range tt.errors {
+					require.Contains(t, cerr.Error(), error)
+				}
+				require.Equal(t, tt.errCode, cerr.StatusCode())
+			} else {
+				require.NoError(t, err)
+			}
+			for path, expect := range tt.expected {
+				b, err := afero.ReadFile(fs, path)
+				require.NoError(t, err)
+				require.Equal(t, expect, string(b))
+			}
+		})
+	}
+}

agent/immortalstreams/backedpipe/backed_pipe.go

@@ -0,0 +1,350 @@
+package backedpipe
+
+import (
+	"context"
+	"io"
+	"sync"
+
+	"golang.org/x/sync/errgroup"
+	"golang.org/x/sync/singleflight"
+	"golang.org/x/xerrors"
+)
+
+var (
+	ErrPipeClosed             = xerrors.New("pipe is closed")
+	ErrPipeAlreadyConnected   = xerrors.New("pipe is already connected")
+	ErrReconnectionInProgress = xerrors.New("reconnection already in progress")
+	ErrReconnectFailed        = xerrors.New("reconnect failed")
+	ErrInvalidSequenceNumber  = xerrors.New("remote sequence number exceeds local sequence")
+	ErrReconnectWriterFailed  = xerrors.New("reconnect writer failed")
+)
+
+// connectionState represents the current state of the BackedPipe connection.
+type connectionState int
+
+const (
+	// connected indicates the pipe is connected and operational.
+	connected connectionState = iota
+	// disconnected indicates the pipe is not connected but not closed.
+	disconnected
+	// reconnecting indicates a reconnection attempt is in progress.
+	reconnecting
+	// closed indicates the pipe is permanently closed.
+	closed
+)
+
+// ErrorEvent represents an error from a reader or writer with connection generation info.
+type ErrorEvent struct {
+	Err        error
+	Component  string // "reader" or "writer"
+	Generation uint64 // connection generation when error occurred
+}
+
+const (
+	// Default buffer capacity used by the writer - 64MB
+	DefaultBufferSize = 64 * 1024 * 1024
+)
+
+// Reconnector is an interface for establishing connections when the BackedPipe needs to reconnect.
+// Implementations should:
+// 1. Establish a new connection to the remote side
+// 2. Exchange sequence numbers with the remote side
+// 3. Return the new connection and the remote's reader sequence number
+//
+// The readerSeqNum parameter is the local reader's current sequence number
+// (total bytes successfully read from the remote). This must be sent to the
+// remote so it can replay its data to us starting from this number.
+//
+// The returned remoteReaderSeqNum should be the remote side's reader sequence
+// number (how many bytes of our outbound data it has successfully read). This
+// informs our writer where to resume (i.e., which bytes to replay to the remote).
+type Reconnector interface {
+	Reconnect(ctx context.Context, readerSeqNum uint64) (conn io.ReadWriteCloser, remoteReaderSeqNum uint64, err error)
+}
+
+// BackedPipe provides a reliable bidirectional byte stream over unreliable network connections.
+// It orchestrates a BackedReader and BackedWriter to provide transparent reconnection
+// and data replay capabilities.
+type BackedPipe struct {
+	ctx         context.Context
+	cancel      context.CancelFunc
+	mu          sync.RWMutex
+	reader      *BackedReader
+	writer      *BackedWriter
+	reconnector Reconnector
+	conn        io.ReadWriteCloser
+
+	// State machine
+	state   connectionState
+	connGen uint64 // Increments on each successful reconnection
+
+	// Unified error handling with generation filtering
+	errChan chan ErrorEvent
+
+	// singleflight group to dedupe concurrent ForceReconnect calls
+	sf singleflight.Group
+
+	// Track first error per generation to avoid duplicate reconnections
+	lastErrorGen uint64
+}
+
+// NewBackedPipe creates a new BackedPipe with default options and the specified reconnector.
+// The pipe starts disconnected and must be connected using Connect().
+func NewBackedPipe(ctx context.Context, reconnector Reconnector) *BackedPipe {
+	pipeCtx, cancel := context.WithCancel(ctx)
+
+	errChan := make(chan ErrorEvent, 1)
+
+	bp := &BackedPipe{
+		ctx:         pipeCtx,
+		cancel:      cancel,
+		reconnector: reconnector,
+		state:       disconnected,
+		connGen:     0, // Start with generation 0
+		errChan:     errChan,
+	}
+
+	// Create reader and writer with typed error channel for generation-aware error reporting
+	bp.reader = NewBackedReader(errChan)
+	bp.writer = NewBackedWriter(DefaultBufferSize, errChan)
+
+	// Start error handler goroutine
+	go bp.handleErrors()
+
+	return bp
+}
+
+// Connect establishes the initial connection using the reconnect function.
+func (bp *BackedPipe) Connect() error {
+	bp.mu.Lock()
+	defer bp.mu.Unlock()
+
+	if bp.state == closed {
+		return ErrPipeClosed
+	}
+
+	if bp.state == connected {
+		return ErrPipeAlreadyConnected
+	}
+
+	// Use internal context for the actual reconnect operation to ensure
+	// Close() reliably cancels any in-flight attempt.
+	return bp.reconnectLocked()
+}
+
+// Read implements io.Reader by delegating to the BackedReader.
+func (bp *BackedPipe) Read(p []byte) (int, error) {
+	return bp.reader.Read(p)
+}
+
+// Write implements io.Writer by delegating to the BackedWriter.
+func (bp *BackedPipe) Write(p []byte) (int, error) {
+	bp.mu.RLock()
+	writer := bp.writer
+	state := bp.state
+	bp.mu.RUnlock()
+
+	if state == closed {
+		return 0, io.EOF
+	}
+
+	return writer.Write(p)
+}
+
+// Close closes the pipe and all underlying connections.
+func (bp *BackedPipe) Close() error {
+	bp.mu.Lock()
+	defer bp.mu.Unlock()
+
+	if bp.state == closed {
+		return nil
+	}
+
+	bp.state = closed
+	bp.cancel() // Cancel main context
+
+	// Close all components in parallel to avoid deadlocks
+	//
+	// IMPORTANT: The connection must be closed first to unblock any
+	// readers or writers that might be holding the mutex on Read/Write
+	var g errgroup.Group
+
+	if bp.conn != nil {
+		conn := bp.conn
+		g.Go(func() error {
+			return conn.Close()
+		})
+		bp.conn = nil
+	}
+
+	if bp.reader != nil {
+		reader := bp.reader
+		g.Go(func() error {
+			return reader.Close()
+		})
+	}
+
+	if bp.writer != nil {
+		writer := bp.writer
+		g.Go(func() error {
+			return writer.Close()
+		})
+	}
+
+	// Wait for all close operations to complete and return any error
+	return g.Wait()
+}
+
+// Connected returns whether the pipe is currently connected.
+func (bp *BackedPipe) Connected() bool {
+	bp.mu.RLock()
+	defer bp.mu.RUnlock()
+	return bp.state == connected && bp.reader.Connected() && bp.writer.Connected()
+}
+
+// reconnectLocked handles the reconnection logic. Must be called with write lock held.
+func (bp *BackedPipe) reconnectLocked() error {
+	if bp.state == reconnecting {
+		return ErrReconnectionInProgress
+	}
+
+	bp.state = reconnecting
+	defer func() {
+		// Only reset to disconnected if we're still in reconnecting state
+		// (successful reconnection will set state to connected)
+		if bp.state == reconnecting {
+			bp.state = disconnected
+		}
+	}()
+
+	// Close existing connection if any
+	if bp.conn != nil {
+		_ = bp.conn.Close()
+		bp.conn = nil
+	}
+
+	// Increment the generation and update both reader and writer.
+	// We do it now to track even the connections that fail during
+	// Reconnect.
+	bp.connGen++
+	bp.reader.SetGeneration(bp.connGen)
+	bp.writer.SetGeneration(bp.connGen)
+
+	// Reconnect reader and writer
+	seqNum := make(chan uint64, 1)
+	newR := make(chan io.Reader, 1)
+
+	go bp.reader.Reconnect(seqNum, newR)
+
+	// Get the precise reader sequence number from the reader while it holds its lock
+	readerSeqNum, ok := <-seqNum
+	if !ok {
+		// Reader was closed during reconnection
+		return ErrReconnectFailed
+	}
+
+	// Perform reconnect using the exact sequence number we just received
+	conn, remoteReaderSeqNum, err := bp.reconnector.Reconnect(bp.ctx, readerSeqNum)
+	if err != nil {
+		// Unblock reader reconnect
+		newR <- nil
+		return ErrReconnectFailed
+	}
+
+	// Provide the new connection to the reader (reader still holds its lock)
+	newR <- conn
+
+	// Replay our outbound data from the remote's reader sequence number
+	writerReconnectErr := bp.writer.Reconnect(remoteReaderSeqNum, conn)
+	if writerReconnectErr != nil {
+		return ErrReconnectWriterFailed
+	}
+
+	// Success - update state
+	bp.conn = conn
+	bp.state = connected
+
+	return nil
+}
+
+// handleErrors listens for connection errors from reader/writer and triggers reconnection.
+// It filters errors from old connections and ensures only the first error per generation
+// triggers reconnection.
+func (bp *BackedPipe) handleErrors() {
+	for {
+		select {
+		case <-bp.ctx.Done():
+			return
+		case errorEvt := <-bp.errChan:
+			bp.handleConnectionError(errorEvt)
+		}
+	}
+}
+
+// handleConnectionError handles errors from either reader or writer components.
+// It filters errors from old connections and ensures only one reconnection per generation.
+func (bp *BackedPipe) handleConnectionError(errorEvt ErrorEvent) {
+	bp.mu.Lock()
+	defer bp.mu.Unlock()
+
+	// Skip if already closed
+	if bp.state == closed {
+		return
+	}
+
+	// Filter errors from old connections (lower generation)
+	if errorEvt.Generation < bp.connGen {
+		return
+	}
+
+	// Skip if not connected (already disconnected or reconnecting)
+	if bp.state != connected {
+		return
+	}
+
+	// Skip if we've already seen an error for this generation
+	if bp.lastErrorGen >= errorEvt.Generation {
+		return
+	}
+
+	// This is the first error for this generation
+	bp.lastErrorGen = errorEvt.Generation
+
+	// Mark as disconnected
+	bp.state = disconnected
+
+	// Try to reconnect using internal context
+	reconnectErr := bp.reconnectLocked()
+
+	if reconnectErr != nil {
+		// Reconnection failed - log or handle as needed
+		// For now, we'll just continue and wait for manual reconnection
+		_ = errorEvt.Err       // Use the original error from the component
+		_ = errorEvt.Component // Component info available for potential logging by higher layers
+	}
+}
+
+// ForceReconnect forces a reconnection attempt immediately.
+// This can be used to force a reconnection if a new connection is established.
+// It prevents duplicate reconnections when called concurrently.
+func (bp *BackedPipe) ForceReconnect() error {
+	// Deduplicate concurrent ForceReconnect calls so only one reconnection
+	// attempt runs at a time from this API. Use the pipe's internal context
+	// to ensure Close() cancels any in-flight attempt.
+	_, err, _ := bp.sf.Do("force-reconnect", func() (interface{}, error) {
+		bp.mu.Lock()
+		defer bp.mu.Unlock()
+
+		if bp.state == closed {
+			return nil, io.EOF
+		}
+
+		// Don't force reconnect if already reconnecting
+		if bp.state == reconnecting {
+			return nil, ErrReconnectionInProgress
+		}
+
+		return nil, bp.reconnectLocked()
+	})
+	return err
+}