Merge branch 'main' into security-patch-train-doc

## Summary

Replaces N per-chat heartbeat goroutines with a single centralized
heartbeat loop that issues one `UPDATE` per 30s interval for all running
chats on a worker.

## Problem

Each running chat spawned a dedicated goroutine that issued an
individual `UPDATE chats SET heartbeat_at = NOW() WHERE id = $1 AND
worker_id = $2 AND status = 'running'` query every 30 seconds. At 10,000
concurrent chats this produces **~333 DB queries/second** just for
heartbeats, plus ~333 `ActivityBumpWorkspace` CTE queries/second from
`trackWorkspaceUsage`.

## Solution

New `UpdateChatHeartbeats` (plural) SQL query replaces the old singular
`UpdateChatHeartbeat`:

```sql
UPDATE chats
SET    heartbeat_at = @now::timestamptz
WHERE  worker_id = @worker_id::uuid
  AND  status = 'running'::chat_status
RETURNING id;
```

A single `heartbeatLoop` goroutine on the `Server`:
1. Ticks every `chatHeartbeatInterval` (30s)
2. Issues one batch UPDATE for all registered chats
3. Detects stolen/completed chats via set-difference (equivalent of old
`rows == 0`)
4. Calls `trackWorkspaceUsage` for surviving chats

`processChat` registers an entry in the heartbeat registry instead of
spawning a goroutine.

## Impact

| Metric | Before (10K chats) | After (10K chats) |
|---|---|---|
| Heartbeat queries/sec | ~333 | ~0.03 (1 per 30s per replica) |
| Heartbeat goroutines | 10,000 | 1 |
| Self-interrupt detection | Per-chat `rows==0` | Batch set-difference |

---

> 🤖 Generated by Coder Agents

<details><summary>Implementation notes</summary>

- Uses `@now` parameter instead of `NOW()` so tests with `quartz.Mock`
can control timestamps.
- `heartbeatEntry` stores `context.CancelCauseFunc` + workspace state
for the centralized loop.
- `recoverStaleChats` is unaffected — it reads `heartbeat_at` which is
still updated.
- The old singular `UpdateChatHeartbeat` is removed entirely.
- `dbauthz` wrapper uses system-level `rbac.ResourceChat` authorization
(same pattern as `AcquireChats`).

</details>

.agents/skills/deep-review/SKILL.md

@@ -111,8 +111,8 @@ Tier 2 file filters:
 
 - **Modernization Reviewer**: one instance per language present in the diff. Filter by extension:
   - Go: `*.go` — reference `.claude/docs/GO.md` before reviewing.
-  - TypeScript: `*.ts` `*.tsx`
-  - React: `*.tsx` `*.jsx`
+  - TypeScript: `*.ts` `*.tsx`: reference `.agents/skills/deep-review/references/typescript.md` before reviewing.
+  - React: `*.tsx` `*.jsx`: reference `.agents/skills/deep-review/references/react.md` before reviewing.
 
   `.tsx` files match both TypeScript and React filters. Spawn both instances when the diff contains `.tsx` changes — TS covers language-level patterns; React covers component and hooks patterns. Before spawning, verify each instance's filter produces a non-empty diff. Skip instances whose filtered diff is empty.
 
@@ -155,9 +155,11 @@ File scope: {filter from step 2}.
 Output file: {REVIEW_DIR}/{role-name}.md
 ```
 
-For the Modernization Reviewer (Go), add after the methodology line:
+For Modernization Reviewer instances, add the language reference after the methodology line:
 
-> Read `.claude/docs/GO.md` as your Go language reference before reviewing.
+- **Go:** `Read .claude/docs/GO.md as your Go language reference before reviewing.`
+- **TypeScript:** `Read .agents/skills/deep-review/references/typescript.md as your TypeScript language reference before reviewing.`
+- **React:** `Read .agents/skills/deep-review/references/react.md as your React language reference before reviewing.`
 
 For re-reviews, append to both Tier 1 and Tier 2 prompts:
 

.agents/skills/deep-review/references/react.md

@@ -0,0 +1,305 @@
+# Modern React (18–19.2) + Compiler 1.0 — Reference
+
+Reference for writing idiomatic React. Covers what changed, what it replaced, and what to reach for. Includes React Compiler patterns — what the compiler handles automatically, what it changes semantically, and how to verify its behavior empirically. Scope: client-side SPA patterns only. Server Components, `use server`, and `use client` directives are framework-specific and omitted. Check the project's React version and compiler config before reaching for newer APIs.
+
+## How modern React thinks differently
+
+**Concurrent rendering** (18): React can now pause, interrupt, and resume renders. This is the foundation everything else builds on. Most existing code "just works," but components that produce side effects during render (mutations, subscriptions, network calls in the render body) are unsafe and will misbehave. Concurrent features are opt-in — they only activate when you use a concurrent API like `startTransition` or `useDeferredValue`.
+
+**Urgent vs. non-urgent updates** (18): The `startTransition` / `useTransition` API introduces a formal split between updates that must feel immediate (typing, clicking) and updates that can be interrupted (filtering a large list, navigating to a new screen). Non-urgent updates yield to urgent ones mid-render. Use this instead of `setTimeout` or manual debounce when you want the UI to stay responsive during expensive re-renders.
+
+**Actions** (19): Async functions passed to `startTransition` are called "Actions." They automatically manage pending state, error handling, and optimistic updates as a unit. The `useActionState` hook and `<form action={fn}>` prop are built on this. The pattern replaces the hand-rolled `isPending/setIsPending` + `try/catch` + `setError` boilerplate that was previously necessary for every data mutation.
+
+**Automatic batching** (18): State updates are now batched everywhere — inside `setTimeout`, `Promise.then`, native event handlers, etc. Previously batching only happened inside React-managed event handlers. If you genuinely need a synchronous flush, use `flushSync`.
+
+**Automatic memoization** (Compiler 1.0): React Compiler is a build-time Babel plugin that automatically inserts memoization into components and hooks. It replaces manual `useMemo`, `useCallback`, and `React.memo` — including conditional memoization and memoization after early returns, which manual APIs cannot express. The compiler only processes components and hooks, not standalone functions. It understands data flow and mutability through its own HIR (High-level Intermediate Representation), so it can memoize more granularly than a human would. Projects adopt it incrementally — typically via path-based Babel overrides or the `"use memo"` directive. Components that violate the Rules of React are silently skipped (no build error), so the automated lint tools that check compiler compatibility matter.
+
+## Replace these patterns
+
+The left column reflects patterns common before React 18/19. Write the right column instead. The "Since" column tells you the minimum React version required.
+
+| Old pattern                                                       | Modern replacement                                                             | Since |
+| ----------------------------------------------------------------- | ------------------------------------------------------------------------------ | ----- |
+| `ReactDOM.render(<App />, el)`                                    | `createRoot(el).render(<App />)`                                               | 18    |
+| `ReactDOM.hydrate(<App />, el)`                                   | `hydrateRoot(el, <App />)`                                                     | 18    |
+| `ReactDOM.unmountComponentAtNode(el)`                             | `root.unmount()`                                                               | 18    |
+| `ReactDOM.findDOMNode(this)`                                      | DOM ref: `const ref = useRef(); ref.current`                                   | 18    |
+| `<Context.Provider value={v}>`                                    | `<Context value={v}>`                                                          | 19    |
+| `React.forwardRef((props, ref) => ...)`                           | `function Comp({ ref, ...props }) { ... }` (ref as a regular prop)             | 19    |
+| String ref `ref="input"` in class components                      | Callback ref or `createRef()`                                                  | 19    |
+| `Heading.propTypes = { ... }`                                     | TypeScript / ES6 type annotations                                              | 19    |
+| `Component.defaultProps = { ... }` on function components         | ES6 default parameters `({ text = 'Hi' })`                                     | 19    |
+| Legacy Context: `contextTypes` + `getChildContext`                | `React.createContext()` + `contextType`                                        | 19    |
+| `import { act } from 'react-dom/test-utils'`                      | `import { act } from 'react'`                                                  | 19    |
+| `import ShallowRenderer from 'react-test-renderer/shallow'`       | `import ShallowRenderer from 'react-shallow-renderer'`                         | 19    |
+| Manual `isPending` state around async calls                       | `const [isPending, startTransition] = useTransition()`                         | 18    |
+| Manual optimistic state + revert logic                            | `useOptimistic(currentValue)`                                                  | 19    |
+| `useEffect` to subscribe to external stores                       | `useSyncExternalStore(subscribe, getSnapshot)`                                 | 18    |
+| Hand-rolled unique ID (counter, random, index)                    | `useId()` — SSR-safe, hydration-safe                                           | 18    |
+| `useEffect` to inject `<title>` or `<meta>` / `react-helmet`      | Render `<title>`, `<meta>`, `<link>` directly in components; React hoists them | 19    |
+| `ReactDOM.useFormState(action, initial)` (Canary name)            | `useActionState(action, initial)`                                              | 19    |
+| `useReducer<React.Reducer<State, Action>>(reducer)`               | `useReducer(reducer)` — infers from the reducer function                       | 19    |
+| `<div ref={current => (instance = current)} />` (implicit return) | `<div ref={current => { instance = current }} />` (explicit block body)        | 19    |
+| `useRef<T>()` with no argument                                    | `useRef<T>(undefined)` or `useRef<T \| null>(null)` — argument is now required | 19    |
+| `MutableRefObject<T>` type annotation                             | `RefObject<T>` — all refs are mutable now; `MutableRefObject` is deprecated    | 19    |
+| `React.createFactory('button')`                                   | `<button />` JSX                                                               | 19    |
+| `useMemo(() => expr, [deps])` in compiled components              | `const val = expr;` — compiler memoizes automatically                          | C 1.0 |
+| `useCallback(fn, [deps])` in compiled components                  | `const fn = () => { ... };` — compiler memoizes automatically                  | C 1.0 |
+| `React.memo(Component)` in compiled components                    | Plain component — compiler skips re-render when props are unchanged            | C 1.0 |
+| `eslint-plugin-react-compiler` (standalone)                       | `eslint-plugin-react-hooks@latest` (compiler rules merged into recommended)    | C 1.0 |
+| `useRef` + `useLayoutEffect` for stable callbacks                 | `useEffectEvent(fn)` — compiler handles both, but `useEffectEvent` is clearer  | 19.2  |
+
+## New capabilities
+
+These enable things that weren't practical before. Reach for them in the described situations.
+
+| What                                                                 | Since   | When to use it                                                                                                                                                                                                                                                                                                                                                                                                                                                              |
+| -------------------------------------------------------------------- | ------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `useTransition()` / `startTransition()`                              | 18      | Mark a state update as non-urgent so React can interrupt it to handle clicks or keystrokes. The `isPending` boolean lets you show a loading indicator without blocking the UI.                                                                                                                                                                                                                                                                                              |
+| `useDeferredValue(value, initialValue?)`                             | 18 / 19 | Defer re-rendering a slow subtree: pass the deferred value as a prop, wrap the expensive child in `memo`. Unlike debounce, uses no fixed timeout — renders as soon as the browser is idle. The `initialValue` arg (19) avoids a flash on first render.                                                                                                                                                                                                                      |
+| `useId()`                                                            | 18      | Generate a stable, SSR-consistent ID for accessibility attributes (`htmlFor`, `aria-describedby`). Do not use for list keys.                                                                                                                                                                                                                                                                                                                                                |
+| `useSyncExternalStore(subscribe, getSnapshot, getServerSnapshot?)`   | 18      | Subscribe to external (non-React) state stores safely under concurrent rendering. Preferred over `useEffect`-based subscriptions in libraries.                                                                                                                                                                                                                                                                                                                              |
+| `useActionState(action, initialState)`                               | 19      | Manage an async mutation: returns `[state, wrappedAction, isPending]`. Handles pending, result, and error state as a unit. Replaces the manual `isPending` + `try/catch` + `setError` pattern.                                                                                                                                                                                                                                                                              |
+| `useOptimistic(currentValue)`                                        | 19      | Show a speculative value while an async Action is in flight. Returns `[optimisticValue, setOptimistic]`. React automatically reverts to `currentValue` when the transition settles.                                                                                                                                                                                                                                                                                         |
+| `use(promiseOrContext)`                                              | 19      | Read a promise or Context value inside a component or custom hook. Unlike hooks, `use` can be called conditionally (after early returns). Promises must come from a cache — do not create them during render.                                                                                                                                                                                                                                                               |
+| `useFormStatus()` (from `react-dom`)                                 | 19      | Read `{ pending, data, method, action }` of the nearest parent `<form>` Action. Works across component boundaries without prop drilling — useful for submit buttons inside design-system components.                                                                                                                                                                                                                                                                        |
+| `useEffectEvent(fn)`                                                 | 19.2    | Extract a non-reactive callback from an effect. The function sees the latest props/state without being listed in deps, and is never stale. Replaces the `useRef`-and-mutate-in-layout-effect workaround for stable event-like callbacks. The compiler has built-in knowledge of this hook and correctly prunes its return value from effect dependency arrays. Both `useEffectEvent` and the old ref workaround compile cleanly; `useEffectEvent` is preferred for clarity. |
+| `<Activity>`                                                         | 19.2    | Hide part of the UI while preserving its state and DOM. React deprioritizes updates to hidden content. Use via framework APIs for route prerendering or tab preservation — not a direct replacement for CSS `visibility`.                                                                                                                                                                                                                                                   |
+| `captureOwnerStack()`                                                | 19.1    | Dev-only API that returns a string showing which components are responsible for rendering the current component (owner stack, not call stack). Useful for custom error overlays. Returns `null` in production.                                                                                                                                                                                                                                                              |
+| `<form action={fn}>`                                                 | 19      | Pass an async function as a form's `action` prop. React handles submission, pending state, and automatic form reset on success. Works with `useActionState` and `useFormStatus`.                                                                                                                                                                                                                                                                                            |
+| Ref cleanup function                                                 | 19      | Return a cleanup function from a ref callback: `ref={el => { ...; return () => cleanup(); }}`. React calls it on unmount. Replaces the pattern of checking `el === null` in the callback.                                                                                                                                                                                                                                                                                   |
+| `<link rel="stylesheet" precedence="default">`                       | 19      | Declare a stylesheet next to the component that needs it. React deduplicates and inserts it in the correct order before revealing Suspense content.                                                                                                                                                                                                                                                                                                                         |
+| `preinit`, `preload`, `prefetchDNS`, `preconnect` (from `react-dom`) | 19      | Imperatively hint the browser to load resources early. Call from render or event handlers. React deduplicates hints across the component tree.                                                                                                                                                                                                                                                                                                                              |
+| React Compiler (`babel-plugin-react-compiler`)                       | C 1.0   | Build-time automatic memoization for components and hooks. Install, add to Babel/Vite pipeline. Projects typically start with path-based overrides to compile a subset of files.                                                                                                                                                                                                                                                                                            |
+| `"use memo"` directive                                               | C 1.0   | Opt a single function into compilation when using `compilationMode: 'annotation'`. Place at the start of the function body. Module-level `"use memo"` at the top of a file compiles all functions in that file.                                                                                                                                                                                                                                                             |
+| `"use no memo"` directive                                            | C 1.0   | Temporary escape hatch — skip compilation for a specific component or hook that causes a runtime regression. Not a permanent solution. Place at the start of the function body.                                                                                                                                                                                                                                                                                             |
+| Compiler-powered ESLint rules                                        | C 1.0   | Rules for purity, refs, set-state-in-render, immutability, etc. now ship in `eslint-plugin-react-hooks` recommended preset. Surface Rules-of-React violations even without the compiler installed. Note: some projects use Biome instead — check project lint config.                                                                                                                                                                                                       |
+
+## Key APIs
+
+### `useTransition` and `startTransition` (18)
+
+`useTransition` returns `[isPending, startTransition]`. Wrap any state update that is not directly tied to the user's current gesture inside `startTransition`. React will render the old UI while computing the new one, and `isPending` is `true` during that window.
+
+In React 19, `startTransition` can accept an async function (an "Action"). React sets `isPending` to `true` for the entire duration of the async work, not just during the synchronous part.
+
+```tsx
+// 18: synchronous transition
+const [isPending, startTransition] = useTransition();
+startTransition(() => setQuery(input));
+
+// 19: async Action — isPending stays true until the await settles
+startTransition(async () => {
+	const err = await updateName(name);
+	if (err) setError(err);
+});
+```
+
+Use `startTransition` (the module-level export) when you cannot use the hook (outside a component, in a router callback, etc.).
+
+### `useDeferredValue` (18 / 19)
+
+Creates a "lagging" copy of a value. Pass it to a memoized, expensive component so that React can render the stale UI while computing the updated one.
+
+```tsx
+// 19: initialValue shows '' on first render; avoids loading flash
+const deferred = useDeferredValue(searchQuery, "");
+return <Results query={deferred} />; // Results wrapped in memo
+```
+
+`deferred !== searchQuery` while the deferred render is in progress — use this to show a "stale" indicator.
+
+### `useActionState` (19)
+
+Replaces the `useState` + `isPending` + `try/catch` + `setError` boilerplate for any async operation that can be retried or submitted as a form.
+
+```tsx
+const [error, submitAction, isPending] = useActionState(
+	async (prevState, formData) => {
+		const err = await updateName(formData.get("name"));
+		if (err) return err; // returned value becomes next state
+		redirect("/profile");
+		return null;
+	},
+	null, // initialState
+);
+
+// Use submitAction as the form's action prop or call it directly
+<form action={submitAction}>
+	<input name="name" />
+	<button disabled={isPending}>Save</button>
+	{error && <p>{error}</p>}
+</form>;
+```
+
+### `useOptimistic` (19)
+
+Shows a speculative value immediately while an async Action is in progress. React automatically reverts to the server-confirmed value when the Action resolves or rejects.
+
+```tsx
+const [optimisticName, setOptimisticName] = useOptimistic(currentName);
+
+const submit = async (formData) => {
+	const newName = formData.get("name");
+	setOptimisticName(newName); // shows immediately
+	await updateName(newName); // reverts if this throws
+};
+```
+
+### `use()` (19)
+
+Unlike hooks, `use` can appear after conditional statements. Two primary uses:
+
+**Reading a promise** (must be stable — from a cache, not created inline):
+
+```tsx
+function Comments({ commentsPromise }) {
+	const comments = use(commentsPromise); // suspends until resolved
+	return comments.map((c) => <p key={c.id}>{c.text}</p>);
+}
+```
+
+**Reading context after an early return** (hooks cannot appear after `return`):
+
+```tsx
+function Heading({ children }) {
+	if (!children) return null;
+	const theme = use(ThemeContext); // valid here; hooks would not be
+	return <h1 style={{ color: theme.color }}>{children}</h1>;
+}
+```
+
+### `useSyncExternalStore` (18)
+
+The correct way for libraries (and app code) to subscribe to non-React state. Prevents tearing under concurrent rendering.
+
+```tsx
+const value = useSyncExternalStore(
+	store.subscribe, // called when store changes
+	store.getSnapshot, // returns current value (must be stable reference if unchanged)
+	store.getServerSnapshot, // optional: for SSR
+);
+```
+
+## Verifying compiler behavior
+
+The compiler is a black box unless you inspect its output. When reviewing code in compiled paths, run the compiler on the specific code to see what it actually does. Do not guess — verify.
+
+**Run the compiler on a code snippet:**
+
+```sh
+cd site && node -e "
+const {transformSync} = require('@babel/core');
+const code = \`<paste component here>\`;
+const diagnostics = [];
+const result = transformSync(code, {
+  plugins: [
+    ['@babel/plugin-syntax-typescript', {isTSX: true}],
+    ['babel-plugin-react-compiler', {
+      logger: {
+        logEvent(_, event) {
+          if (event.kind === 'CompileError' || event.kind === 'CompileSkip') {
+            diagnostics.push(event.detail?.toString?.()?.substring(0, 200));
+          }
+        },
+      },
+    }],
+  ],
+  filename: 'test.tsx',
+});
+console.log('Compiled:', result.code.includes('_c('));
+if (diagnostics.length) console.log('Diagnostics:', diagnostics);
+console.log(result.code);
+"
+```
+
+**Reading compiled output:**
+
+- `const $ = _c(N)` — allocates N memoization cache slots.
+- `if ($[n] !== dep)` — cache invalidation guard. Re-computes when `dep` changes (referential equality).
+- `if ($[n] === Symbol.for("react.memo_cache_sentinel"))` — one-time initialization. Runs once on first render, cached forever after. This is how the compiler handles expressions with no reactive dependencies.
+- `_temp` functions — pure callbacks the compiler hoisted out of the component body.
+
+**Check all compiled files at once:**
+
+```sh
+cd site && pnpm run lint:compiler
+```
+
+This runs the compiler on every file in the compiled paths and reports CompileError / CompileSkip diagnostics. Zero diagnostics means all functions compiled cleanly.
+
+**What the compiler catches vs. what it does not:**
+
+The compiler emits `CompileError` for mutations of props, state, or hook arguments during render, and for `ref.current` access during render. The project's lint pipeline catches these automatically — do not flag them in review.
+
+The compiler does **not** flag impure function calls during render (`Math.random()`, `Date.now()`, `new Date()`). Instead it silently memoizes them with a sentinel guard, freezing the value after first render. This changes semantics without any diagnostic. Verify suspicious calls by running the compiler and checking for sentinel guards in the output.
+
+## Pitfalls
+
+Things that are easy to get wrong even when you know the modern API exists. Check your output against these.
+
+**Effects run twice in development with StrictMode.** React 18 intentionally mounts → unmounts → remounts every component in dev to surface effects that are not resilient to remounting. This is not a bug. If an effect breaks on the second mount, it is missing a cleanup function. Write `return () => cleanup()` from every effect that sets up a subscription, timer, or external resource.
+
+**Concurrent rendering can call render multiple times.** The render function (component body) may be called more than once before React commits to the DOM. Side effects (mutations, subscriptions, logging) in the render body will run multiple times. Move them into `useEffect` or event handlers.
+
+**Do not create promises during render and pass them to `use()`.** A new promise is created every render, causing an infinite suspend-retry loop. Create the promise outside the component (module level), or use a caching library (SWR, React Query, `cache()` from React) to stabilize it.
+
+**`useOptimistic` reverts automatically — do not fight it.** The optimistic value is a presentation layer only. When the Action settles, React replaces it with the real `currentValue` you passed in. Do not try to sync optimistic state back to your real state; let React handle the revert.
+
+**`flushSync` opts out of automatic batching.** If third-party code or a browser API (e.g. `ResizeObserver`) calls `setState` and you need synchronous DOM flushing, wrap with `flushSync(() => setState(...))`. This is a last resort; prefer letting React batch.
+
+**`forwardRef` still works in React 19 but will be deprecated.** Function components accept `ref` as a plain prop now. New code should use the prop directly. Existing `forwardRef` wrappers continue to work without changes; migrate when convenient.
+
+**`<Activity>` does not unmount.** Content inside a hidden `<Activity>` boundary stays mounted. Effects keep running. Use it for preserving scroll position or form state, not for preventing expensive mounts — use lazy loading for that.
+
+**TypeScript: implicit returns from ref callbacks are now type errors.** In React 19, returning anything other than a cleanup function (or nothing) from a ref callback is rejected by the TypeScript types. The most common case is arrow-function refs that implicitly return the DOM node:
+
+```tsx
+// Error in React 19 types:
+<div ref={el => (instance = el)} />
+
+// Fix — use a block body:
+<div ref={el => { instance = el; }} />
+```
+
+**TypeScript: `useRef` now requires an argument.** `useRef<T>()` with no argument is a type error. Pass `undefined` for mutable refs or `null` for DOM refs you initialize on mount: `useRef<T>(undefined)` / `useRef<HTMLDivElement | null>(null)`.
+
+**`useId` output format changed across versions.** React 18 produced `:r0:`. React 19.1 changed it to `«r0»`. React 19.2 changed it again to `_r0`. Do not parse or depend on the specific format — treat it as an opaque string.
+
+**`useFormStatus` reads the nearest parent `<form>` with a function `action`.** It does not reflect native HTML form submissions — only React Actions. A submit button that is a sibling of `<form>` (rather than a descendant) will not see the form's status.
+
+**Context as a provider (`<Context>`) requires React 19; `<Context.Provider>` still works.** Do not use `<Context>` shorthand in a codebase that needs to support React 18. The two forms can coexist during migration.
+
+**Compiler freezes impure expressions silently.** `Math.random()`, `Date.now()`, `new Date()`, and `window.innerWidth` in a component body all compile without diagnostics. The compiler wraps them in a sentinel guard (`Symbol.for("react.memo_cache_sentinel")`) that runs the expression once and caches the result forever. The value never updates on re-render. Fix: move to a `useState` initializer (`useState(() => Math.random())`), `useEffect`, or event handler.
+
+**Component granularity affects compiler optimization.** When one pattern in a component causes a `CompileError` (e.g., a necessary `ref.current` read during render), the compiler skips the **entire** component. If the rest of the component would benefit from compilation, extract the non-compilable pattern into a small child component. This keeps the parent compiled.
+
+**The compiler only memoizes components and hooks.** Standalone utility functions (even expensive ones called during render) are not compiled. If a utility function is truly expensive, it still needs its own caching strategy outside of React (e.g., a module-level cache, `WeakMap`, etc.).
+
+**Changing memoization can shift `useEffect` firing.** A value that was unstable before compilation may become stable after, causing an effect that depended on it to fire less often. Conversely, future compiler changes may alter memoization granularity. Effects that use memoized values as dependencies should be resilient to these changes — they should be true synchronization effects, not "run this when X changes" hacks.
+
+## Behavioral changes that affect code
+
+- **Automatic batching** (18): State updates in `setTimeout`, `Promise.then`, `addEventListener` callbacks, etc. are now batched into a single re-render. Previously only React synthetic event handlers were batched. Code that relied on unbatched updates (reading DOM synchronously after each `setState`) must use `flushSync`.
+
+- **StrictMode double-invoke** (18): In development, every component is mounted → unmounted → remounted with the previous state. Every effect runs cleanup → setup twice on initial mount. `useMemo` and `useCallback` also double-invoke their functions. Production behavior is unchanged. If a test or component breaks under this, the component had a latent cleanup bug.
+
+- **StrictMode ref double-invoke** (19): In development, ref callbacks are also invoked twice on mount (attach → detach → attach). Return a cleanup function from the ref callback to handle detach correctly.
+
+- **StrictMode memoization reuse** (19): During the second pass of double-rendering, `useMemo` and `useCallback` now reuse the cached result from the first pass instead of calling the function again. Components that are already StrictMode-compatible should not notice a difference.
+
+- **Suspense fallback commits immediately** (19): When a component suspends, React now commits the nearest `<Suspense>` fallback without waiting for sibling trees to finish rendering. After the fallback is shown, React "pre-warms" suspended siblings in the background. This makes fallbacks appear faster but changes the order of rendering work.
+
+- **Error re-throwing removed** (19): Errors that are not caught by an Error Boundary are now reported to `window.reportError` (not re-thrown). Errors caught by an Error Boundary go to `console.error` once. If your production monitoring relied on the re-thrown error, add handlers to `createRoot`: `createRoot(el, { onUncaughtError, onCaughtError })`.
+
+- **Transitions in `popstate` are synchronous** (19): Browser back/forward navigation triggers synchronous transition flushing. This ensures the URL and UI update together atomically during history navigation.
+
+- **`useEffect` from discrete events flushes synchronously** (18): Effects triggered by a click or keydown (discrete events) are now flushed synchronously before the browser paints, consistent with `useLayoutEffect` for those cases.
+
+- **Hydration mismatches treated as errors** (18 / improved in 19): Text content mismatches between server HTML and client render revert to client rendering up to the nearest `<Suspense>` boundary. React 19 logs a single diff instead of multiple warnings, making mismatches much easier to diagnose.
+
+- **New JSX transform required** (19): The automatic JSX runtime introduced in 2020 (`react/jsx-runtime`) is now mandatory. The classic transform (which required `import React from 'react'` in every file) is no longer supported. Most toolchains have already shipped the new transform; check your Babel or TypeScript config if you see warnings.
+
+- **UMD builds removed** (19): React no longer ships UMD bundles. Load via npm and a bundler, or use an ESM CDN (`import React from "https://esm.sh/react@19"`).
+
+- **React Compiler automatic memoization** (Compiler 1.0): Build-time Babel plugin that inserts memoization into components and hooks. Components that follow the Rules of React are automatically memoized; components that violate them are silently skipped (no build error, no runtime change). The compiler can memoize conditionally and after early returns — things impossible with manual `useMemo`/`useCallback`. Works with React 17+ via `react-compiler-runtime`; best with React 19+. Projects adopt incrementally via path-based Babel overrides, `compilationMode: 'annotation'`, or the `"use memo"` / `"use no memo"` directives. Check the project's Vite/Babel config to know which paths are compiled. Compiled components show a "Memo ✨" badge in React DevTools.

.agents/skills/deep-review/references/typescript.md

@@ -0,0 +1,199 @@
+# Modern TypeScript (5.0–6.0 RC) — Reference
+
+Reference for writing idiomatic TypeScript. Covers what changed, what it replaced, and what to reach for. Respect the project's minimum TypeScript version: don't emit features from a version newer than what the project targets. Check `package.json` and `tsconfig.json` before writing code.
+
+## How modern TypeScript thinks differently
+
+The 5.x era resolves years of module system ambiguity and cleans house on legacy options. Three themes dominate:
+
+**Module semantics are explicit.** `--verbatimModuleSyntax` (5.0) makes import/export intent visible in source: type imports must carry `type`, value imports stay. Combined with `--module preserve` or `--moduleResolution bundler`, the compiler now accurately models what bundlers and modern runtimes actually do. `import defer` (5.9) extends the model to deferred evaluation.
+
+**Resource lifetimes are first-class.** `using` and `await using` (5.2) provide deterministic cleanup without `try/finally`. Any object implementing `Symbol.dispose` participates. `DisposableStack` handles ad-hoc multi-resource cleanup in functions where creating a full class is overkill.
+
+**Inference is smarter about what it knows.** Inferred type predicates (5.5) let `.filter(x => x !== undefined)` produce `T[]` instead of `(T | undefined)[]` automatically. `NoInfer<T>` (5.4) gives library authors precise control over which parameters drive inference. Narrowing now survives closures after last assignment, constant indexed accesses, and `switch (true)` patterns.
+
+**TypeScript 6.0 is a transition release toward 7.0** (the Go-native port). It turns years of soft deprecations into errors and changes several defaults. Most impactful: `types` defaults to `[]` (must list `@types` packages explicitly), `rootDir` defaults to `.`, `strict` defaults to `true`, `module` defaults to `esnext`. Projects relying on implicit behavior need explicit config. Check the deprecations section before upgrading.
+
+## Replace these patterns
+
+The left column reflects patterns still common before TypeScript 5.x. Write the right column instead. The "Since" column tells you the minimum TypeScript version required.
+
+| Old pattern                                                                                                      | Modern replacement                                                                                                           | Since                            |
+| ---------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------- | -------------------------------- | ------ |
+| `--experimentalDecorators` + legacy decorator signatures                                                         | Standard decorators (TC39): `function dec(target, context: ClassMethodDecoratorContext)` — no flag needed                    | 5.0                              |
+| Requiring callers to add `as const` at call sites                                                                | `<const T extends HasNames>(arg: T)` — `const` modifier on type parameter                                                    | 5.0                              |
+| `--importsNotUsedAsValues` + `--preserveValueImports`                                                            | `--verbatimModuleSyntax`                                                                                                     | 5.0                              |
+| `import { Foo } from "..."` when `Foo` is only used as a type                                                    | `import { type Foo } from "..."` or `import type { Foo } from "..."`                                                         | 5.0                              |
+| `"extends": "@tsconfig/strictest/tsconfig.json"` chain                                                           | `"extends": ["@tsconfig/strictest/tsconfig.json", "./tsconfig.base.json"]` (array form)                                      | 5.0                              |
+| `try { ... } finally { resource.close(); resource.delete(); }`                                                   | `using resource = acquireResource()` — calls `[Symbol.dispose]()` automatically                                              | 5.2                              |
+| `try { ... } finally { await resource.close() }`                                                                 | `await using resource = acquireAsyncResource()`                                                                              | 5.2                              |
+| Ad-hoc cleanup with multiple `try/finally` blocks                                                                | `using cleanup = new DisposableStack(); cleanup.defer(() => ...)`                                                            | 5.2                              |
+| `import data from "./data.json" assert { type: "json" }`                                                         | `import data from "./data.json" with { type: "json" }`                                                                       | 5.3                              |
+| `.filter(Boolean)` or `.filter(x => !!x)` to remove nulls                                                        | `.filter(x => x !== undefined)` or `.filter(x => x !== null)` (infers type predicate)                                        | 5.5                              |
+| Extra phantom type param to block inference bleed: `<C extends string, D extends C>`                             | `NoInfer<C>` on the parameter you don't want to drive inference                                                              | 5.4                              |
+| `/** @typedef {import("./types").Foo} Foo */` in JS files                                                        | `/** @import { Foo } from "./types" */` (JSDoc `@import` tag)                                                                | 5.5                              |
+| `myArray.reverse()` mutating in place                                                                            | `myArray.toReversed()` (returns new array)                                                                                   | 5.2                              |
+| `myArray.sort(cmp)` mutating in place                                                                            | `myArray.toSorted(cmp)` (returns new array)                                                                                  | 5.2                              |
+| `const copy = [...arr]; copy[i] = v`                                                                             | `arr.with(i, v)` (returns new array)                                                                                         | 5.2                              |
+| Manual `has`/`get`/`set` pattern on `Map`                                                                        | `map.getOrInsert(key, defaultValue)` or `getOrInsertComputed(key, fn)`                                                       | 6.0 RC                           |
+| `new RegExp(str.replace(/[.\*+?^${}()                                                                            | [\]\\]/g, '\\$&'))`                                                                                                          | `new RegExp(RegExp.escape(str))` | 6.0 RC |
+| `--moduleResolution node` (node10)                                                                               | `--moduleResolution nodenext` (Node.js) or `--moduleResolution bundler` (bundlers/Bun)                                       | 6.0 RC                           |
+| `"baseUrl": "./src"` + `"@app/*": ["app/*"]` in paths                                                            | Remove `baseUrl`; use `"@app/*": ["./src/app/*"]` in paths directly                                                          | 6.0 RC                           |
+| `module Foo { export const x = 1; }`                                                                             | `namespace Foo { export const x = 1; }`                                                                                      | 6.0 RC                           |
+| `export * from "..."` when all re-exported members are types                                                     | `export type * from "..."` (or `export type * as ns from "..."`)                                                             | 5.0                              |
+| `function f(): undefined { return undefined; }` — explicit return required in `: undefined`-returning function   | Remove the `return` entirely; `undefined`-returning functions no longer require any return statement                         | 5.1                              |
+| Manual type predicate annotation on a simple arrow: `(x: T \| undefined): x is T => x !== undefined`             | Remove the annotation; TypeScript infers `x is T` from `!== null/undefined` and `instanceof` checks automatically            | 5.5                              |
+| `const val = obj[key]; if (typeof val === "string") { use(val); }` — extract to const to narrow indexed access   | `if (typeof obj[key] === "string") { obj[key].toUpperCase(); }` directly — both `obj` and `key` must be effectively constant | 5.5                              |
+| Copy narrowed `let`/param to a `const`, or restructure code to escape stale closure narrowing after reassignment | Remove the copy; narrowing survives into closures created after the last assignment to the variable                          | 5.4                              |
+| `(arr as string[]).filter(...)` or restructure to avoid "not callable" errors on `string[] \| number[]`          | Call `.filter`, `.find`, `.some`, `.every`, `.reduce` directly on union-of-array types                                       | 5.2                              |
+| `if`/`else` chain used to work around lack of narrowing inside a `switch (true)` body                            | `switch (true)` — each `case` condition now narrows the tested variable in its clause                                        | 5.3                              |
+
+## New capabilities
+
+These enable things that weren't practical before. Reach for them in the described situations.
+
+| What                                            | Since  | When to use it                                                                                                                                                                                                                                |
+| ----------------------------------------------- | ------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `using` / `await using` declarations            | 5.2    | Any resource needing deterministic cleanup (file handles, DB connections, locks, event listeners). Object must implement `Symbol.dispose` / `Symbol.asyncDispose`.                                                                            |
+| `DisposableStack` / `AsyncDisposableStack`      | 5.2    | Ad-hoc multi-resource cleanup without creating a class. Call `.defer(fn)` right after acquiring each resource. Stack disposes in LIFO order.                                                                                                  |
+| `const` modifier on type parameters             | 5.0    | Force `const`-like (literal/readonly tuple) inference at call sites without requiring callers to write `as const`. Constraint must use `readonly` arrays.                                                                                     |
+| Decorator metadata (`Symbol.metadata`)          | 5.2    | Attach and read per-class metadata from decorators via `context.metadata`. Retrieved as `MyClass[Symbol.metadata]`. Requires `Symbol.metadata ??= Symbol(...)` polyfill.                                                                      |
+| `NoInfer<T>` utility type                       | 5.4    | Prevent a parameter from contributing inference candidates for `T`. Use when one argument should be the "source of truth" and others should only be checked against it.                                                                       |
+| Inferred type predicates                        | 5.5    | Filter callbacks that test for `!== null` or `instanceof` now automatically produce a type predicate. `Array.prototype.filter` then narrows the result array type.                                                                            |
+| `--isolatedDeclarations`                        | 5.5    | Require explicit return types on exported declarations. Unlocks parallel declaration emit by external tooling (esbuild, oxc, etc.) without needing a full type-checker pass.                                                                  |
+| `${configDir}` in tsconfig paths                | 5.5    | Anchor `typeRoots`, `paths`, `outDir`, etc. in a shared base tsconfig to the _consuming_ project's directory, not the shared file's location.                                                                                                 |
+| Always-truthy/nullish check errors              | 5.6    | Catches regex literals in `if`, arrow functions as comparators, `?? 100` on non-nullable left side, misplaced parentheses. No API to call; existing bugs now surface as errors.                                                               |
+| Iterator helper methods (`IteratorObject`)      | 5.6    | Built-in iterators from `Map`, `Set`, generators, etc. now have `.map()`, `.filter()`, `.take()`, `.drop()`, `.flatMap()`, `.toArray()`, `.reduce()`, etc. Use `Iterator.from(iterable)` to wrap any iterable.                                |
+| `--noUncheckedSideEffectImports`                | 5.6    | Error when a side-effect import (`import "..."`) resolves to nothing. Catches typos in polyfill or CSS imports.                                                                                                                               |
+| `--noCheck`                                     | 5.6    | Skip type checking entirely during emit. Useful for separating "fast emit" from "thorough check" pipeline stages, especially with `--isolatedDeclarations`.                                                                                   |
+| `--rewriteRelativeImportExtensions`             | 5.7    | Rewrite `.ts`→`.js`, `.tsx`→`.jsx`, `.mts`→`.mjs`, `.cts`→`.cjs` in relative imports during emit. Required when writing `.ts` imports for Node.js strip-types mode and still needing `.js` output for library distribution.                   |
+| `--erasableSyntaxOnly`                          | 5.8    | Error on constructs that can't be type-stripped by Node.js `--experimental-strip-types`: `enum`, `namespace` with code, parameter properties, `import =` aliases.                                                                             |
+| `require()` of ESM under `--module nodenext`    | 5.8    | Node.js 22+ allows CJS to `require()` ESM files (no top-level `await`). TypeScript now allows this under `nodenext` without error.                                                                                                            |
+| `import defer * as ns from "..."`               | 5.9    | Defer module _evaluation_ (not loading) until first property access. Module is loaded and verified at import time; side-effects are delayed. Only works with `--module preserve` or `esnext`.                                                 |
+| `Set` algebra methods                           | 5.5    | Non-mutating: `union`, `intersection`, `difference`, `symmetricDifference` → new `Set`. Predicate: `isSubsetOf`, `isSupersetOf`, `isDisjointFrom` → `boolean`. Requires `esnext` or `es2025` lib.                                             |
+| `Object.groupBy` / `Map.groupBy`                | 5.4    | Group an iterable into buckets by key function. Return type has all keys as optional (not every key is guaranteed present). Requires `esnext` or `es2024`+ lib.                                                                               |
+| `Temporal` API types                            | 6.0 RC | `Temporal.Now`, `Temporal.Instant`, `Temporal.PlainDate`, etc. Available under `esnext` or `esnext.temporal` lib. Usable in runtimes that already ship it (V8 118+, SpiderMonkey, etc.).                                                      |
+| `@satisfies` in JSDoc                           | 5.0    | Validates that a JS expression satisfies a type without widening it — the TS `satisfies` operator for `.js` files. Write `/** @satisfies {MyType} */` above the declaration or inline on a parenthesized expression.                          |
+| `@overload` in JSDoc                            | 5.0    | Declare multiple call signatures for a JS function. Each JSDoc comment tagged `@overload` is treated as a distinct overload; the final JSDoc comment (without `@overload`) describes the implementation signature.                            |
+| Getter/setter with completely unrelated types   | 5.1    | `get style(): CSSStyleDeclaration` and `set style(v: string)` can now have fully unrelated types, provided both have explicit type annotations. Previously the getter type was required to be a subtype of the setter type.                   |
+| `instanceof` narrowing via `Symbol.hasInstance` | 5.3    | When a class defines `static [Symbol.hasInstance](val: unknown): val is T`, the `instanceof` operator now narrows to the predicate type `T`, not the class type itself. Useful when the runtime check and the structural type differ.         |
+| Regex literal syntax checking                   | 5.5    | TypeScript validates regex literal syntax: malformed groups, nonexistent backreferences, named capture mismatches, and features not available at the current `--target`. No API needed; existing latent bugs surface as errors automatically. |
+| `--build` continues past intermediate errors    | 5.6    | `tsc --build` no longer stops at the first failing project. All projects are built and errors reported together. Use `--stopOnBuildErrors` to restore the old stop-on-first-error behavior. Useful for monorepos during upgrades.             |
+| `--module node18`                               | 5.8    | Stable `--module` flag for Node.js 18 semantics: disallows `require()` of ESM (unlike `nodenext`) and still allows import assertions. Use when pinned to Node 18 and not ready for `nodenext` behavior changes.                               |
+| `--module node20`                               | 5.9    | Stable `--module` flag for Node.js 20 semantics: permits `require()` of ESM, rejects import assertions. Implies `--target es2023` (unlike `nodenext`, which floats to `esnext`).                                                              |
+
+## Key APIs
+
+### `Disposable` / `AsyncDisposable` / stacks (5.2)
+
+Global types provided by TypeScript's lib (requires `esnext.disposable` or `esnext` in `lib`):
+
+- `Disposable` — `{ [Symbol.dispose](): void }`
+- `AsyncDisposable` — `{ [Symbol.asyncDispose](): PromiseLike<void> }`
+- `DisposableStack` — `defer(fn)`, `use(resource)`, `adopt(value, disposeFn)`, `move()`. Is itself `Disposable`.
+- `AsyncDisposableStack` — async equivalent. Is itself `AsyncDisposable`.
+- `SuppressedError` — thrown when both the scope body and a `[Symbol.dispose]` throw. `.error` holds the dispose-phase error; `.suppressed` holds the original error.
+
+Polyfill the symbols in older runtimes:
+
+```ts
+Symbol.dispose ??= Symbol("Symbol.dispose");
+Symbol.asyncDispose ??= Symbol("Symbol.asyncDispose");
+```
+
+### Decorator context types (5.0)
+
+Each decorator kind receives a typed context object as its second parameter:
+
+- `ClassDecoratorContext`
+- `ClassMethodDecoratorContext`
+- `ClassGetterDecoratorContext`
+- `ClassSetterDecoratorContext`
+- `ClassFieldDecoratorContext`
+- `ClassAccessorDecoratorContext`
+
+All context objects have `.name`, `.kind`, `.static`, `.private`, and `.metadata`. Method/getter/setter/accessor contexts also have `.addInitializer(fn)` for running code at construction time.
+
+### `IteratorObject` (5.6)
+
+`IteratorObject<T, TReturn, TNext>` is the new type for built-in iterable iterators. Key methods: `map`, `filter`, `take`, `drop`, `flatMap`, `forEach`, `reduce`, `some`, `every`, `find`, `toArray`. Not the same as the pre-existing structural `Iterator<T>` protocol.
+
+- Generators produce `Generator<T>` which extends `IteratorObject`.
+- `Map.prototype.entries()` returns `MapIterator<[K, V]>`, `Set.prototype.values()` returns `SetIterator<T>`, etc.
+- `Iterator.from(iterable)` converts any `Iterable` to an `IteratorObject`.
+- `AsyncIteratorObject` exists for async parity.
+- `--strictBuiltinIteratorReturn` (new `--strict`-mode flag in 5.6) makes the return type of `BuiltinIteratorReturn` be `undefined` instead of `any`, catching unchecked `done` access.
+
+### Array copying methods (5.2)
+
+Declared on `Array`, `ReadonlyArray`, and all `TypedArray` types. Use these instead of the mutating variants when you need to preserve the original:
+
+| Mutating                           | Non-mutating copy                     |
+| ---------------------------------- | ------------------------------------- |
+| `arr.sort(cmp)`                    | `arr.toSorted(cmp)`                   |
+| `arr.reverse()`                    | `arr.toReversed()`                    |
+| `arr.splice(start, del, ...items)` | `arr.toSpliced(start, del, ...items)` |
+| `arr[i] = v`                       | `arr.with(i, v)`                      |
+
+## Pitfalls
+
+Things easy to get wrong even when you know the modern API exists. Check your output against these.
+
+**tsconfig defaults changed hard in 6.0.** `types: []` means no `@types/*` packages load implicitly. If you see floods of "cannot find name 'process'" or "cannot find module 'fs'" after upgrading to 6.0, add `"types": ["node"]` (or whatever you need) to `compilerOptions`. `rootDir: "."` means a project with source in `src/` will emit to `dist/src/` instead of `dist/` — add `"rootDir": "./src"` explicitly. `strict: true` by default means projects with loose code see new errors.
+
+**`using` requires a runtime polyfill on older runtimes.** `Symbol.dispose` and `Symbol.asyncDispose` don't exist before Node.js 18.x / Chrome 120. Add the two-line polyfill at your entry point. `DisposableStack` and `AsyncDisposableStack` need a more substantial polyfill (e.g. from `@microsoft/using-polyfill`).
+
+**`using` disposes in LIFO order.** Resources declared later in a scope are disposed first. Declare in the order you want reversed cleanup (acquisition order). `DisposableStack.defer` also runs in LIFO order.
+
+**Inferred type predicates have if-and-only-if semantics.** `x => !!x` does NOT infer `x is NonNullable<T>` because `0`, `""`, and `false` are falsy but not absent. TypeScript correctly refuses the predicate. Use `x => x !== undefined` or `x => x !== null` for precise null/undefined filters. If a predicate isn't being inferred, the false branch is probably ambiguous.
+
+**`--verbatimModuleSyntax` breaks CJS `require` emit.** Under this flag ESM `import`/`export` is emitted verbatim. You cannot produce `require()` calls from standard `import` syntax. For CJS output you must use `import foo = require("foo")` and `export = { ... }` syntax explicitly.
+
+**`NoInfer<T>` doesn't prevent `T` from being resolved, only from being contributed at that position.** Other parameters can still infer `T`. It means "don't use me as an inference candidate", not "block `T` from being resolved".
+
+**`--isolatedDeclarations` requires explicit return types on all exports.** Exported arrow functions, function declarations, and class methods all need annotations if their return type isn't trivially inferrable from a literal or type assertion. Editor quick-fixes can add them automatically.
+
+**Standard decorators are incompatible with `--experimentalDecorators`.** Different type signatures, metadata model, and emit. A decorator written for one will not work with the other. `--emitDecoratorMetadata` is not supported with standard decorators. Don't mix the two systems in one project.
+
+**`import defer` does not downlevel.** TypeScript does not transform `import defer` to polyfill-compatible code. The module is still _loaded_ eagerly (must exist); only _evaluation_ is deferred. Only use it under `--module preserve` or `esnext` with a runtime or bundler that supports it.
+
+**`--erasableSyntaxOnly` prohibits parameter properties.** `constructor(public x: number)` is not allowed. Expand to an explicit field declaration plus assignment in the constructor body.
+
+**Closure narrowing is invalidated if the variable is assigned anywhere in a nested function.** TypeScript cannot know when a nested function will run, so any assignment to a `let`/param inside a nested function — even a no-op like `value = value` — invalidates narrowing for all closures in the outer scope. Only the outer "no further assignments after this point" pattern is safe.
+
+**Constant indexed access narrowing requires both `obj` and `key` to be unmodified between the check and the use.** If either is a `let` that could be reassigned, TypeScript will not narrow `obj[key]`. Extract the value to a `const` in that case.
+
+**`switch (true)` narrowing does not carry across fall-through cases.** In a `switch (true)`, each `case` condition narrows independently. A variable narrowed in `case typeof x === "string":` that falls through to the next case will have its narrowing widened by the next condition, not accumulated from the previous one.
+
+**`const` type parameter modifier falls back when constraint is mutable.** `<const T extends string[]>(args: T)` falls back to `string[]` because `readonly ["a", "b"]` isn't assignable to `string[]`. Use `<const T extends readonly string[]>` for arrays.
+
+**`assert` import syntax errors under `--module nodenext` since 5.8.** Any remaining `import x from "..." assert { ... }` must be updated to `import x from "..." with { ... }`.
+
+**`Array.prototype.filter(x => x !== null)` now narrows to non-null (5.5).** This is almost always correct, but if you intentionally needed the nullable type downstream, add an explicit annotation: `const items: (T | null)[] = arr.filter(x => x !== null)`.
+
+## Behavioral changes that affect code
+
+- **All enums are union enums** (5.0): Every enum member gets its own literal type. Out-of-domain literal assignment to an enum type now errors. Cross-enum assignment between enums with identical names but differing values now errors.
+- **Relational operators no longer allow implicit string/number coercions** (5.0): `ns > 4` where `ns: number | string` is a type error. Use `+ns > 4` to explicitly coerce.
+- **`--module`/`--moduleResolution` must agree on node flavor** (5.2): Mixing `--module nodenext` with `--moduleResolution bundler` is an error. Use `--module nodenext` alone or `--module esnext --moduleResolution bundler`.
+- **Deprecations from 5.0 become hard errors in 5.5**: `--importsNotUsedAsValues`, `--preserveValueImports`, `--target ES3`, `--out`, and several others are fully removed in 5.5. They can no longer be specified, even with `"ignoreDeprecations": "5.0"`. Migrate to `--verbatimModuleSyntax` for the import flags.
+- **Type-only imports conflicting with local values** (5.4): Under `--isolatedModules`, `import { Foo } from "..."` where a local `let Foo` also exists now errors. Use `import type { Foo }` or `import { type Foo }`.
+- **Reference directives no longer synthesized or preserved in declaration emit** (5.5): `/// <reference types="node" />` TypeScript used to add automatically is no longer emitted. User-written directives are dropped unless they carry `preserve="true"`. Update library `tsconfig.json` if you relied on this.
+- **`.mts` files never emit CJS; `.cts` files never emit ESM** (5.6): Regardless of `--module` setting. Previously the extension was ignored in some modes.
+- **JSON imports under `--module nodenext` require `with { type: "json" }`** (5.7): `import data from "./config.json"` without the attribute is now a type error.
+- **`TypedArray`s are now generic** (5.7): `Uint8Array` is `Uint8Array<TArrayBuffer extends ArrayBufferLike = ArrayBufferLike>`. Code passing `Buffer` (from `@types/node`) to typed-array parameters may see new errors. Update `@types/node` to a version that matches.
+- **`import assert { ... }` is an error under `--module nodenext`** (5.8): Node.js 22 dropped support for the old syntax. Use `with { ... }`.
+- **`types` defaults to `[]` in 6.0**: All implicit `@types/*` loading stops. Add an explicit `"types": ["node"]` or the array will remain empty. Using `"types": ["*"]` restores the 5.x behavior.
+- **`rootDir` defaults to `.` (the tsconfig directory) in 6.0**: Previously inferred from the common ancestor of all source files. Projects with `"include": ["./src"]` and no explicit `rootDir` will now emit into `dist/src/` instead of `dist/`. Add `"rootDir": "./src"` to fix.
+- **`strict` defaults to `true` in 6.0**: Projects that were implicitly not strict will see new errors. Set `"strict": false` explicitly if you're not ready to fix them.
+- **`--baseUrl` deprecated in 6.0** and no longer acts as a module resolution root. Add explicit prefixes to your `paths` entries instead.
+- **`--moduleResolution node` (node10) deprecated in 6.0**: Removed in 7.0. Migrate to `nodenext` or `bundler`.
+- **`amd`, `umd`, `systemjs`, `none` module targets deprecated in 6.0**: Removed in 7.0. Migrate to a bundler.
+- **`--outFile` removed in 6.0**: Use a bundler (esbuild, Rollup, Webpack, etc.).
+- **`module Foo { }` syntax removed in 6.0**: Rename all such declarations to `namespace Foo { }`.
+- **`--esModuleInterop false` and `--allowSyntheticDefaultImports false` removed in 6.0**: Safe interop is now always on. Default imports from CJS modules (`import express from "express"`) are always valid.
+- **Explicit `typeRoots` disables upward `node_modules/@types` fallback** (5.1): When `typeRoots` is specified and a lookup fails in those directories, TypeScript no longer walks parent directories for `@types`. If you relied on the fallback, add `"./node_modules/@types"` explicitly to your `typeRoots` array.
+- **`super.` on instance field properties is a type error** (5.3): Calling `super.foo()` where `foo` is a class field (arrow function assigned in the constructor) rather than a prototype method now errors. Instance fields don't exist on the prototype; `super.field` is `undefined` at runtime.
+- **`--build` always emits `.tsbuildinfo`** (5.6): Previously only written when `--incremental` or `--composite` was set. Now written unconditionally in any `--build` invocation. Update `.gitignore` or CI artifact management if needed.
+- **`.mts`/`.cts` extensions and `package.json` `"type"` respected in all module modes** (5.6): Format-specific extensions and the `"type"` field inside `node_modules` are now honored regardless of `--module` setting (except `amd`, `umd`, `system`). A `.mts` file will never emit CJS output even under `--module commonjs`.
+- **Granular return expression checking** (5.8): Each branch of a conditional expression (`cond ? a : b`) directly inside a `return` statement is now checked individually against the declared return type. Previously an `any`-typed branch could silently suppress type errors in the other branch.

.github/dependabot.yaml

@@ -31,7 +31,8 @@ updates:
         patterns:
           - "golang.org/x/*"
     ignore:
-      # Ignore patch updates for all dependencies
+      # Patch updates are handled by the security-patch-prs workflow so this
+      # lane stays focused on broader dependency updates.
       - dependency-name: "*"
         update-types:
           - version-update:semver-patch
@@ -56,7 +57,7 @@ updates:
     labels: []
     ignore:
       # We need to coordinate terraform updates with the version hardcoded in
-      # our Go code.
+      # our Go code. These are handled by the security-patch-prs workflow.
       - dependency-name: "terraform"
 
   - package-ecosystem: "npm"
@@ -82,9 +83,6 @@ updates:
       mui:
         patterns:
           - "@mui*"
-      radix:
-        patterns:
-          - "@radix-ui/*"
       react:
         patterns:
           - "react"
@@ -120,11 +118,11 @@ updates:
       interval: "weekly"
     commit-message:
       prefix: "chore"
+    labels: []
     groups:
       coder-modules:
         patterns:
           - "coder/*/coder"
-    labels: []
     ignore:
       - dependency-name: "*"
         update-types:

.github/workflows/ci.yaml

@@ -35,7 +35,7 @@ jobs:
       tailnet-integration: ${{ steps.filter.outputs.tailnet-integration }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -157,7 +157,7 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -204,7 +204,7 @@ jobs:
 
       # Needed for helm chart linting
       - name: Install helm
-        uses: azure/setup-helm@1a275c3b69536ee54be43f2070a358922e12c8d4 # v4.3.1
+        uses: azure/setup-helm@dda3372f752e03dde6b3237bc9431cdc2f7a02a2 # v5.0.0
         with:
           version: v3.9.2
         continue-on-error: true
@@ -247,7 +247,7 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -272,7 +272,7 @@ jobs:
     if: ${{ !cancelled() }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -327,7 +327,7 @@ jobs:
     timeout-minutes: 20
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -379,7 +379,7 @@ jobs:
           - windows-2022
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -575,7 +575,7 @@ jobs:
     timeout-minutes: 25
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -637,7 +637,7 @@ jobs:
     timeout-minutes: 25
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -709,7 +709,7 @@ jobs:
     timeout-minutes: 20
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -736,7 +736,7 @@ jobs:
     timeout-minutes: 20
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -769,7 +769,7 @@ jobs:
     name: ${{ matrix.variant.name }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -849,7 +849,7 @@ jobs:
     if: needs.changes.outputs.site == 'true' || needs.changes.outputs.ci == 'true'
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -870,7 +870,7 @@ jobs:
       # the check to pass. This is desired in PRs, but not in mainline.
       - name: Publish to Chromatic (non-mainline)
         if: github.ref != 'refs/heads/main' && github.repository_owner == 'coder'
-        uses: chromaui/action@07791f8243f4cb2698bf4d00426baf4b2d1cb7e0 # v13.3.5
+        uses: chromaui/action@f191a0224b10e1a38b2091cefb7b7a2337009116 # v16.0.0
         env:
           NODE_OPTIONS: "--max_old_space_size=4096"
           STORYBOOK: true
@@ -902,7 +902,7 @@ jobs:
       # infinitely "in progress" in mainline unless we re-review each build.
       - name: Publish to Chromatic (mainline)
         if: github.ref == 'refs/heads/main' && github.repository_owner == 'coder'
-        uses: chromaui/action@07791f8243f4cb2698bf4d00426baf4b2d1cb7e0 # v13.3.5
+        uses: chromaui/action@f191a0224b10e1a38b2091cefb7b7a2337009116 # v16.0.0
         env:
           NODE_OPTIONS: "--max_old_space_size=4096"
           STORYBOOK: true
@@ -930,7 +930,7 @@ jobs:
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -1005,7 +1005,7 @@ jobs:
     if: always()
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -1043,7 +1043,7 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -1097,7 +1097,7 @@ jobs:
       IMAGE: ghcr.io/coder/coder-preview:${{ steps.build-docker.outputs.tag }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -1479,7 +1479,7 @@ jobs:
     if: needs.changes.outputs.db == 'true' || needs.changes.outputs.ci == 'true' || github.ref == 'refs/heads/main'
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/dependabot.yaml

@@ -23,7 +23,7 @@ jobs:
     steps:
       - name: Dependabot metadata
         id: metadata
-        uses: dependabot/fetch-metadata@21025c705c08248db411dc16f3619e6b5f9ea21a # v2.5.0
+        uses: dependabot/fetch-metadata@ffa630c65fa7e0ecfa0625b5ceda64399aea1b36 # v3.0.0
         with:
           github-token: "${{ secrets.GITHUB_TOKEN }}"
 

.github/workflows/deploy.yaml

@@ -36,7 +36,7 @@ jobs:
       verdict: ${{ steps.check.outputs.verdict }} # DEPLOY or NOOP
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -65,7 +65,7 @@ jobs:
       packages: write # to retag image as dogfood
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -142,7 +142,7 @@ jobs:
     needs: deploy
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/docker-base.yaml

@@ -38,7 +38,7 @@ jobs:
     if: github.repository_owner == 'coder'
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/dogfood.yaml

@@ -26,7 +26,7 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-4' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -125,7 +125,7 @@ jobs:
       id-token: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/linear-release.yaml

@@ -5,8 +5,6 @@ on:
     branches:
       - main
       - "release/2.[0-9]+"
-  release:
-    types: [published]
 
 permissions:
   contents: read
@@ -29,9 +27,9 @@ jobs:
 
       - name: Detect next release version
         id: version
-        # Find the highest release/2.X branch (exact pattern, no suffixes like
-        # release/2.31_hotfix) and derive the next minor version for the release
-        # currently in development on main.
+        # Find the highest release/2.X branch (exact pattern, no suffixes
+        # like release/2.31_hotfix) and derive the next minor version for
+        # the release currently in development on main.
         run: |
           LATEST_MINOR=$(git branch -r | grep -E '^\s*origin/release/2\.[0-9]+$' | \
             sed 's/.*release\/2\.//' | sort -n | tail -1)
@@ -40,8 +38,10 @@ jobs:
             echo "skip=true" >> "$GITHUB_OUTPUT"
             exit 0
           fi
-          echo "version=2.$((LATEST_MINOR + 1))" >> "$GITHUB_OUTPUT"
+          NEXT="2.$((LATEST_MINOR + 1))"
+          echo "version=$NEXT" >> "$GITHUB_OUTPUT"
           echo "skip=false" >> "$GITHUB_OUTPUT"
+          echo "Detected next release: $NEXT"
 
       - name: Sync issues
         id: sync
@@ -51,6 +51,7 @@ jobs:
           access_key: ${{ secrets.LINEAR_ACCESS_KEY }}
           command: sync
           version: ${{ steps.version.outputs.version }}
+          name: ${{ steps.version.outputs.version }}
           timeout: 300
 
   sync-release-branch:
@@ -76,6 +77,7 @@ jobs:
           access_key: ${{ secrets.LINEAR_ACCESS_KEY }}
           command: sync
           version: ${{ steps.version.outputs.version }}
+          name: ${{ steps.version.outputs.version }}
           timeout: 300
 
   code-freeze:
@@ -106,38 +108,3 @@ jobs:
           version: ${{ steps.version.outputs.version }}
           timeout: 300
 
-  complete:
-    name: Complete Linear release
-    if: github.event_name == 'release'
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
-        with:
-          persist-credentials: false
-
-      - name: Extract release version
-        id: version
-        # Strip "v" prefix and patch: "v2.31.0" -> "2.31". Also detect whether
-        # this is a minor release (v*.*.0) — patch releases (v2.31.1, v2.31.2,
-        # ...) are grouped into the same Linear release and must not re-complete
-        # it after it has already shipped.
-        run: |
-          VERSION=$(echo "$TAG" | sed 's/^v//' | cut -d. -f1,2)
-          echo "version=$VERSION" >> "$GITHUB_OUTPUT"
-          if [[ "$TAG" =~ ^v[0-9]+\.[0-9]+\.0$ ]]; then
-            echo "is_minor=true" >> "$GITHUB_OUTPUT"
-          else
-            echo "is_minor=false" >> "$GITHUB_OUTPUT"
-          fi
-        env:
-          TAG: ${{ github.event.release.tag_name }}
-
-      - name: Complete release
-        id: complete
-        if: steps.version.outputs.is_minor == 'true'
-        uses: linear/linear-release-action@755d50b5adb7dd42b976ee9334952745d62ceb2d # v0.6.0
-        with:
-          access_key: ${{ secrets.LINEAR_ACCESS_KEY }}
-          command: complete
-          version: ${{ steps.version.outputs.version }}
-          timeout: 300

.github/workflows/nightly-gauntlet.yaml

@@ -28,7 +28,7 @@ jobs:
           - windows-2022
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

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

@@ -15,7 +15,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         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@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/pr-deploy.yaml

@@ -39,7 +39,7 @@ jobs:
       PR_OPEN: ${{ steps.check_pr.outputs.pr_open }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -76,7 +76,7 @@ jobs:
     runs-on: "ubuntu-latest"
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -184,7 +184,7 @@ jobs:
       pull-requests: write # needed for commenting on PRs
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -228,7 +228,7 @@ jobs:
       CODER_IMAGE_TAG: ${{ needs.get_info.outputs.CODER_IMAGE_TAG }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -288,7 +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@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/release-validation.yaml

@@ -14,7 +14,7 @@ jobs:
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/release.yaml

@@ -9,6 +9,7 @@ on:
         options:
           - mainline
           - stable
+          - rc
       release_notes:
         description: Release notes for the publishing the release. This is required to create a release.
       dry_run:
@@ -80,7 +81,7 @@ jobs:
       version: ${{ steps.version.outputs.version }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -119,9 +120,19 @@ jobs:
             exit 1
           fi
 
-          # 2.10.2 -> release/2.10
+          # Derive the release branch from the version tag.
+          # Standard: 2.10.2 -> release/2.10
+          # RC:       2.32.0-rc.0 -> release/2.32-rc.0
           version="$(./scripts/version.sh)"
-          release_branch=release/${version%.*}
+          if [[ "$version" == *-rc.* ]]; then
+            # Extract major.minor and rc suffix from e.g. 2.32.0-rc.0
+            base_version="${version%%-rc.*}"   # 2.32.0
+            major_minor="${base_version%.*}"   # 2.32
+            rc_suffix="${version##*-rc.}"      # 0
+            release_branch="release/${major_minor}-rc.${rc_suffix}"
+          else
+            release_branch=release/${version%.*}
+          fi
           branch_contains_tag=$(git branch --remotes --contains "${GITHUB_REF}" --list "*/${release_branch}" --format='%(refname)')
           if [[ -z "${branch_contains_tag}" ]]; then
             echo "Ref tag must exist in a branch named ${release_branch} when creating a release, did you use scripts/release.sh?"
@@ -531,6 +542,9 @@ jobs:
           if [[ $CODER_RELEASE_CHANNEL == "stable" ]]; then
             publish_args+=(--stable)
           fi
+          if [[ $CODER_RELEASE_CHANNEL == "rc" ]]; then
+            publish_args+=(--rc)
+          fi
           if [[ $CODER_DRY_RUN == *t* ]]; then
             publish_args+=(--dry-run)
           fi
@@ -563,6 +577,35 @@ jobs:
           VERSION: ${{ steps.version.outputs.version }}
           CREATED_LATEST_TAG: ${{ steps.build_docker.outputs.created_latest_tag }}
 
+      # Mark the Linear release as shipped.
+      - name: Extract Linear release version
+        if: ${{ !inputs.dry_run }}
+        id: linear_version
+        run: |
+          # Skip RC releases — they must not complete the Linear release.
+          if [[ "$VERSION" == *-rc* ]]; then
+            echo "RC release (${VERSION}), skipping Linear release completion."
+            echo "skip=true" >> "$GITHUB_OUTPUT"
+            exit 0
+          fi
+          # Strip patch to get the Linear release version (e.g. 2.32.0 -> 2.32).
+          linear_version=$(echo "$VERSION" | cut -d. -f1,2)
+          echo "version=$linear_version" >> "$GITHUB_OUTPUT"
+          echo "skip=false" >> "$GITHUB_OUTPUT"
+          echo "Completing Linear release ${linear_version}"
+        env:
+          VERSION: ${{ steps.version.outputs.version }}
+
+      - name: Complete Linear release
+        if: ${{ !inputs.dry_run && steps.linear_version.outputs.skip != 'true' }}
+        continue-on-error: true
+        uses: linear/linear-release-action@755d50b5adb7dd42b976ee9334952745d62ceb2d # v0.6.0
+        with:
+          access_key: ${{ secrets.LINEAR_ACCESS_KEY }}
+          command: complete
+          version: ${{ steps.linear_version.outputs.version }}
+          timeout: 300
+
       - name: Authenticate to Google Cloud
         uses: google-github-actions/auth@7c6bc770dae815cd3e89ee6cdf493a5fab2cc093 # v3.0.0
         with:
@@ -614,7 +657,7 @@ jobs:
           retention-days: 7
 
       - name: Send repository-dispatch event
-        if: ${{ !inputs.dry_run }}
+        if: ${{ !inputs.dry_run && inputs.release_channel != 'rc' }}
         uses: peter-evans/repository-dispatch@28959ce8df70de7be546dd1250a005dd32156697 # v4.0.1
         with:
           token: ${{ secrets.CDRCI_GITHUB_TOKEN }}
@@ -630,7 +673,7 @@ jobs:
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -702,11 +745,11 @@ jobs:
     name: Publish to winget-pkgs
     runs-on: windows-latest
     needs: release
-    if: ${{ !inputs.dry_run }}
+    if: ${{ !inputs.dry_run && inputs.release_channel != 'rc' }}
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/scorecard.yml

@@ -20,7 +20,7 @@ jobs:
 
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -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@5d4e8d1aca955e8d8589aabd499c5cae939e33c7 # v3.29.5
+        uses: github/codeql-action/upload-sarif@c10b8064de6f491fea524254123dbe5e09572f13 # v3.29.5
         with:
           sarif_file: results.sarif

.github/workflows/security-backport.yaml

@@ -0,0 +1,354 @@
+name: security-backport
+
+on:
+  pull_request_target:
+    types:
+      - labeled
+      - unlabeled
+      - closed
+  workflow_dispatch:
+    inputs:
+      pull_request:
+        description: Pull request number to backport.
+        required: true
+        type: string
+
+permissions:
+  contents: write
+  pull-requests: write
+  issues: write
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || inputs.pull_request }}
+  cancel-in-progress: false
+
+env:
+  LATEST_BRANCH: release/2.31
+  STABLE_BRANCH: release/2.30
+  STABLE_1_BRANCH: release/2.29
+
+jobs:
+  label-policy:
+    if: github.event_name == 'pull_request_target'
+    runs-on: ubuntu-latest
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        with:
+          egress-policy: audit
+
+      - name: Apply security backport label policy
+        shell: bash
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -euo pipefail
+
+          pr_number="$(jq -r '.pull_request.number' "${GITHUB_EVENT_PATH}")"
+          pr_json="$(gh pr view "${pr_number}" --json number,title,url,baseRefName,labels)"
+
+          PR_JSON="${pr_json}" \
+          python3 - <<'PY'
+          import json
+          import os
+          import subprocess
+          import sys
+
+          pr = json.loads(os.environ["PR_JSON"])
+
+          pr_number = pr["number"]
+          labels = [label["name"] for label in pr.get("labels", [])]
+
+          def has(label: str) -> bool:
+              return label in labels
+
+          def ensure_label(label: str) -> None:
+              if not has(label):
+                  subprocess.run(
+                      ["gh", "pr", "edit", str(pr_number), "--add-label", label],
+                      check=False,
+                  )
+
+          def remove_label(label: str) -> None:
+              if has(label):
+                  subprocess.run(
+                      ["gh", "pr", "edit", str(pr_number), "--remove-label", label],
+                      check=False,
+                  )
+
+          def comment(body: str) -> None:
+              subprocess.run(
+                  ["gh", "pr", "comment", str(pr_number), "--body", body],
+                  check=True,
+              )
+
+          if not has("security:patch"):
+              remove_label("status:needs-severity")
+              sys.exit(0)
+
+          severity_labels = [
+              label
+              for label in ("severity:medium", "severity:high", "severity:critical")
+              if has(label)
+          ]
+          if len(severity_labels) == 0:
+              ensure_label("status:needs-severity")
+              comment(
+                  "This PR is labeled `security:patch` but is missing a severity "
+                  "label. Add one of `severity:medium`, `severity:high`, or "
+                  "`severity:critical` before backport automation can proceed."
+              )
+              sys.exit(0)
+
+          if len(severity_labels) > 1:
+              comment(
+                  "This PR has multiple severity labels. Keep exactly one of "
+                  "`severity:medium`, `severity:high`, or `severity:critical`."
+              )
+              sys.exit(1)
+
+          remove_label("status:needs-severity")
+
+          target_labels = [
+              label
+              for label in ("backport:stable", "backport:stable-1")
+              if has(label)
+          ]
+          has_none = has("backport:none")
+          if has_none and target_labels:
+              comment(
+                  "`backport:none` cannot be combined with other backport labels. "
+                  "Remove `backport:none` or remove the explicit backport targets."
+              )
+              sys.exit(1)
+
+          if not has_none and not target_labels:
+              ensure_label("backport:stable")
+              ensure_label("backport:stable-1")
+              comment(
+                  "Applied default backport labels `backport:stable` and "
+                  "`backport:stable-1` for a qualifying security patch."
+              )
+          PY
+
+  backport:
+    if: >
+      github.event_name == 'workflow_dispatch' ||
+      (
+        github.event_name == 'pull_request_target' &&
+        github.event.pull_request.merged == true
+      )
+    runs-on: ubuntu-latest
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        with:
+          egress-policy: audit
+
+      - name: Checkout
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        with:
+          fetch-depth: 0
+          persist-credentials: false
+
+      - name: Resolve PR metadata
+        id: metadata
+        shell: bash
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          INPUT_PR_NUMBER: ${{ inputs.pull_request }}
+          LATEST_BRANCH: ${{ env.LATEST_BRANCH }}
+          STABLE_BRANCH: ${{ env.STABLE_BRANCH }}
+          STABLE_1_BRANCH: ${{ env.STABLE_1_BRANCH }}
+        run: |
+          set -euo pipefail
+
+          if [ "${GITHUB_EVENT_NAME}" = "workflow_dispatch" ]; then
+            pr_number="${INPUT_PR_NUMBER}"
+          else
+            pr_number="$(jq -r '.pull_request.number' "${GITHUB_EVENT_PATH}")"
+          fi
+
+          case "${pr_number}" in
+            ''|*[!0-9]*)
+              echo "A valid pull request number is required."
+              exit 1
+              ;;
+          esac
+
+          pr_json="$(gh pr view "${pr_number}" --json number,title,url,mergeCommit,baseRefName,labels,mergedAt,author)"
+
+          PR_JSON="${pr_json}" \
+          python3 - <<'PY'
+          import json
+          import os
+          import sys
+
+          pr = json.loads(os.environ["PR_JSON"])
+          github_output = os.environ["GITHUB_OUTPUT"]
+
+          labels = [label["name"] for label in pr.get("labels", [])]
+          if "security:patch" not in labels:
+              print("Not a security patch PR; skipping.")
+              sys.exit(0)
+
+          severity_labels = [
+              label
+              for label in ("severity:medium", "severity:high", "severity:critical")
+              if label in labels
+          ]
+          if len(severity_labels) != 1:
+              raise SystemExit(
+                  "Merged security patch PR must have exactly one severity label."
+              )
+
+          if not pr.get("mergedAt"):
+              raise SystemExit(f"PR #{pr['number']} is not merged.")
+
+          if "backport:none" in labels:
+              target_pairs = []
+          else:
+              mapping = {
+                  "backport:stable": os.environ["STABLE_BRANCH"],
+                  "backport:stable-1": os.environ["STABLE_1_BRANCH"],
+              }
+              target_pairs = []
+              for label_name, branch in mapping.items():
+                  if label_name in labels and branch and branch != pr["baseRefName"]:
+                      target_pairs.append({"label": label_name, "branch": branch})
+
+          with open(github_output, "a", encoding="utf-8") as f:
+              f.write(f"pr_number={pr['number']}\n")
+              f.write(f"merge_sha={pr['mergeCommit']['oid']}\n")
+              f.write(f"title={pr['title']}\n")
+              f.write(f"url={pr['url']}\n")
+              f.write(f"author={pr['author']['login']}\n")
+              f.write(f"severity_label={severity_labels[0]}\n")
+              f.write(f"target_pairs={json.dumps(target_pairs)}\n")
+          PY
+
+      - name: Backport to release branches
+        if: ${{ steps.metadata.outputs.target_pairs != '[]' }}
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          PR_NUMBER: ${{ steps.metadata.outputs.pr_number }}
+          MERGE_SHA: ${{ steps.metadata.outputs.merge_sha }}
+          PR_TITLE: ${{ steps.metadata.outputs.title }}
+          PR_URL: ${{ steps.metadata.outputs.url }}
+          PR_AUTHOR: ${{ steps.metadata.outputs.author }}
+          SEVERITY_LABEL: ${{ steps.metadata.outputs.severity_label }}
+          TARGET_PAIRS: ${{ steps.metadata.outputs.target_pairs }}
+        run: |
+          set -euo pipefail
+
+          git config user.name "github-actions[bot]"
+          git config user.email "41898282+github-actions[bot]@users.noreply.github.com"
+
+          git remote set-url origin "https://x-access-token:${GH_TOKEN}@github.com/${GITHUB_REPOSITORY}.git"
+          git fetch origin --prune
+
+          merge_parent_count="$(git rev-list --parents -n 1 "${MERGE_SHA}" | awk '{print NF-1}')"
+
+          failures=()
+          successes=()
+
+          while IFS=$'\t' read -r backport_label target_branch; do
+            [ -n "${target_branch}" ] || continue
+
+            safe_branch_name="${target_branch//\//-}"
+            head_branch="backport/${safe_branch_name}/pr-${PR_NUMBER}"
+
+            existing_pr="$(gh pr list \
+              --base "${target_branch}" \
+              --head "${head_branch}" \
+              --state all \
+              --json number,url \
+              --jq '.[0]')"
+            if [ -n "${existing_pr}" ] && [ "${existing_pr}" != "null" ]; then
+              pr_url="$(printf '%s' "${existing_pr}" | jq -r '.url')"
+              successes+=("${target_branch}:existing:${pr_url}")
+              continue
+            fi
+
+            git checkout -B "${head_branch}" "origin/${target_branch}"
+
+            if [ "${merge_parent_count}" -gt 1 ]; then
+              cherry_pick_args=(-m 1 "${MERGE_SHA}")
+            else
+              cherry_pick_args=("${MERGE_SHA}")
+            fi
+
+            if ! git cherry-pick -x "${cherry_pick_args[@]}"; then
+              git cherry-pick --abort || true
+              gh pr edit "${PR_NUMBER}" --add-label "backport:conflict" || true
+              gh pr comment "${PR_NUMBER}" --body \
+                "Automatic backport to \`${target_branch}\` conflicted. The original author or release manager should resolve it manually."
+              failures+=("${target_branch}:cherry-pick failed")
+              continue
+            fi
+
+            git push --force-with-lease origin "${head_branch}"
+
+            body_file="$(mktemp)"
+            printf '%s\n' \
+              "Automated backport of [#${PR_NUMBER}](${PR_URL})." \
+              "" \
+              "- Source PR: #${PR_NUMBER}" \
+              "- Source commit: ${MERGE_SHA}" \
+              "- Target branch: ${target_branch}" \
+              "- Severity: ${SEVERITY_LABEL}" \
+              > "${body_file}"
+
+            pr_url="$(gh pr create \
+              --base "${target_branch}" \
+              --head "${head_branch}" \
+              --title "${PR_TITLE} (backport to ${target_branch})" \
+              --body-file "${body_file}")"
+
+            backport_pr_number="$(gh pr list \
+              --base "${target_branch}" \
+              --head "${head_branch}" \
+              --state open \
+              --json number \
+              --jq '.[0].number')"
+
+            gh pr edit "${backport_pr_number}" \
+              --add-label "security:patch" \
+              --add-label "${SEVERITY_LABEL}" \
+              --add-label "${backport_label}" || true
+
+            successes+=("${target_branch}:created:${pr_url}")
+          done < <(
+            python3 - <<'PY'
+            import json
+            import os
+            for pair in json.loads(os.environ["TARGET_PAIRS"]):
+                print(f"{pair['label']}\t{pair['branch']}")
+            PY
+          )
+
+          summary_file="$(mktemp)"
+          {
+            echo "## Security backport summary"
+            echo
+            if [ "${#successes[@]}" -gt 0 ]; then
+              echo "### Created or existing"
+              for entry in "${successes[@]}"; do
+                echo "- ${entry}"
+              done
+              echo
+            fi
+            if [ "${#failures[@]}" -gt 0 ]; then
+              echo "### Failures"
+              for entry in "${failures[@]}"; do
+                echo "- ${entry}"
+              done
+            fi
+          } | tee -a "${GITHUB_STEP_SUMMARY}" > "${summary_file}"
+
+          gh pr comment "${PR_NUMBER}" --body-file "${summary_file}"
+
+          if [ "${#failures[@]}" -gt 0 ]; then
+            printf 'Backport failures:\n%s\n' "${failures[@]}" >&2
+            exit 1
+          fi

.github/workflows/security-patch-prs.yaml

@@ -0,0 +1,214 @@
+name: security-patch-prs
+
+on:
+  workflow_dispatch:
+  schedule:
+    - cron: "0 3 * * 1-5"
+
+permissions:
+  contents: write
+  pull-requests: write
+
+jobs:
+  patch:
+    strategy:
+      fail-fast: false
+      matrix:
+        lane:
+          - gomod
+          - terraform
+    runs-on: ubuntu-latest
+    steps:
+      - name: Harden Runner
+        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        with:
+          egress-policy: audit
+
+      - name: Checkout
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        with:
+          persist-credentials: false
+
+      - name: Setup Go
+        uses: ./.github/actions/setup-go
+
+      - name: Patch Go dependencies
+        if: matrix.lane == 'gomod'
+        shell: bash
+        run: |
+          set -euo pipefail
+
+          go get -u=patch ./...
+          go mod tidy
+
+          # Guardrail: do not auto-edit replace directives.
+          if git diff --unified=0 -- go.mod | grep -E '^[+-]replace '; then
+            echo "Refusing to auto-edit go.mod replace directives"
+            exit 1
+          fi
+
+          # Guardrail: only go.mod / go.sum may change.
+          extra="$(git diff --name-only | grep -Ev '^(go\.mod|go\.sum)$' || true)"
+          test -z "$extra" || { echo "Unexpected files changed:"; echo "$extra"; exit 1; }
+
+      - name: Patch bundled Terraform
+        if: matrix.lane == 'terraform'
+        shell: bash
+        run: |
+          set -euo pipefail
+
+          current="$(
+            grep -oE 'NewVersion\("[0-9]+\.[0-9]+\.[0-9]+"\)' \
+              provisioner/terraform/install.go \
+            | head -1 \
+            | grep -oE '[0-9]+\.[0-9]+\.[0-9]+'
+          )"
+
+          series="$(echo "$current" | cut -d. -f1,2)"
+
+          latest="$(
+            curl -fsSL https://releases.hashicorp.com/terraform/index.json \
+            | jq -r --arg series "$series" '
+                .versions
+                | keys[]
+                | select(startswith($series + "."))
+              ' \
+            | sort -V \
+            | tail -1
+          )"
+
+          test -n "$latest"
+          [ "$latest" != "$current" ] || exit 0
+
+          CURRENT_TERRAFORM_VERSION="$current" \
+          LATEST_TERRAFORM_VERSION="$latest" \
+          python3 - <<'PY'
+          from pathlib import Path
+          import os
+
+          current = os.environ["CURRENT_TERRAFORM_VERSION"]
+          latest = os.environ["LATEST_TERRAFORM_VERSION"]
+
+          updates = {
+              "scripts/Dockerfile.base": (
+                  f"terraform/{current}/",
+                  f"terraform/{latest}/",
+              ),
+              "provisioner/terraform/install.go": (
+                  f'NewVersion("{current}")',
+                  f'NewVersion("{latest}")',
+              ),
+              "install.sh": (
+                  f'TERRAFORM_VERSION="{current}"',
+                  f'TERRAFORM_VERSION="{latest}"',
+              ),
+          }
+
+          for path_str, (before, after) in updates.items():
+              path = Path(path_str)
+              content = path.read_text()
+              if before not in content:
+                  raise SystemExit(f"did not find expected text in {path_str}: {before}")
+              path.write_text(content.replace(before, after))
+          PY
+
+          # Guardrail: only the Terraform-version files may change.
+          extra="$(git diff --name-only | grep -Ev '^(scripts/Dockerfile.base|provisioner/terraform/install.go|install.sh)$' || true)"
+          test -z "$extra" || { echo "Unexpected files changed:"; echo "$extra"; exit 1; }
+
+      - name: Validate Go dependency patch
+        if: matrix.lane == 'gomod'
+        shell: bash
+        run: |
+          set -euo pipefail
+          go test ./...
+
+      - name: Validate Terraform patch
+        if: matrix.lane == 'terraform'
+        shell: bash
+        run: |
+          set -euo pipefail
+          go test ./provisioner/terraform/...
+          docker build -f scripts/Dockerfile.base .
+
+      - name: Skip PR creation when there are no changes
+        id: changes
+        shell: bash
+        run: |
+          set -euo pipefail
+
+          if git diff --quiet; then
+            echo "has_changes=false" >> "${GITHUB_OUTPUT}"
+          else
+            echo "has_changes=true" >> "${GITHUB_OUTPUT}"
+          fi
+
+      - name: Commit changes
+        if: steps.changes.outputs.has_changes == 'true'
+        shell: bash
+        run: |
+          set -euo pipefail
+
+          git config user.name "github-actions[bot]"
+          git config user.email "41898282+github-actions[bot]@users.noreply.github.com"
+
+          git checkout -B "secpatch/${{ matrix.lane }}"
+          git add -A
+          git commit -m "security: patch ${{ matrix.lane }}"
+
+      - name: Push branch
+        if: steps.changes.outputs.has_changes == 'true'
+        shell: bash
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -euo pipefail
+
+          git push --force-with-lease \
+            "https://x-access-token:${GH_TOKEN}@github.com/${{ github.repository }}.git" \
+            "HEAD:refs/heads/secpatch/${{ matrix.lane }}"
+
+      - name: Create or update PR
+        if: steps.changes.outputs.has_changes == 'true'
+        shell: bash
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -euo pipefail
+
+          branch="secpatch/${{ matrix.lane }}"
+          title="security: patch ${{ matrix.lane }}"
+          body="$(cat <<'EOF'
+          Automated security patch PR for `${{ matrix.lane }}`.
+
+          Scope:
+          - gomod: patch-level Go dependency updates only
+          - terraform: bundled Terraform patch updates only
+
+          Guardrails:
+          - no application-code edits
+          - no auto-editing of go.mod replace directives
+          - CI must pass
+          EOF
+          )"
+
+          existing_pr="$(gh pr list --head "${branch}" --base main --json number --jq '.[0].number // empty')"
+          if [[ -n "${existing_pr}" ]]; then
+            gh pr edit "${existing_pr}" \
+              --title "${title}" \
+              --body "${body}"
+            pr_number="${existing_pr}"
+          else
+            gh pr create \
+              --base main \
+              --head "${branch}" \
+              --title "${title}" \
+              --body "${body}"
+            pr_number="$(gh pr list --head "${branch}" --base main --json number --jq '.[0].number // empty')"
+          fi
+
+          for label in security dependencies automated-pr; do
+            if gh label list --json name --jq '.[].name' | grep -Fxq "${label}"; then
+              gh pr edit "${pr_number}" --add-label "${label}"
+            fi
+          done

.github/workflows/security.yaml

@@ -27,7 +27,7 @@ jobs:
     runs-on: ${{ github.repository_owner == 'coder' && 'depot-ubuntu-22.04-8' || 'ubuntu-latest' }}
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -40,7 +40,7 @@ jobs:
         uses: ./.github/actions/setup-go
 
       - name: Initialize CodeQL
-        uses: github/codeql-action/init@5d4e8d1aca955e8d8589aabd499c5cae939e33c7 # v3.29.5
+        uses: github/codeql-action/init@c10b8064de6f491fea524254123dbe5e09572f13 # v3.29.5
         with:
           languages: go, javascript
 
@@ -50,7 +50,7 @@ jobs:
           rm Makefile
 
       - name: Perform CodeQL Analysis
-        uses: github/codeql-action/analyze@5d4e8d1aca955e8d8589aabd499c5cae939e33c7 # v3.29.5
+        uses: github/codeql-action/analyze@c10b8064de6f491fea524254123dbe5e09572f13 # v3.29.5
 
       - name: Send Slack notification on failure
         if: ${{ failure() }}

.github/workflows/stale.yaml

@@ -18,7 +18,7 @@ jobs:
       pull-requests: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -96,7 +96,7 @@ jobs:
       contents: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -120,7 +120,7 @@ jobs:
       actions: write
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 

.github/workflows/weekly-docs.yaml

@@ -21,7 +21,7 @@ jobs:
       pull-requests: write # required to post PR review comments by the action
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@fa2e9d605c4eeb9fcad4c99c224cee0c6c7f3594 # v2.16.0
+        uses: step-security/harden-runner@fe104658747b27e96e4f7e80cd0a94068e53901d # v2.16.1
         with:
           egress-policy: audit
 
@@ -49,8 +49,10 @@ jobs:
       # TODO: Remove this workaround once action-linkspector sets
       # package-manager-cache: false in its internal setup-node step.
       # See: https://github.com/UmbrellaDocs/action-linkspector/issues/54
-      - name: Enable corepack
-        run: corepack enable pnpm
+      - name: Enable corepack and create pnpm store
+        run: |
+          corepack enable pnpm
+          mkdir -p "$(pnpm store path --silent)"
 
       - name: Check Markdown links
         uses: umbrelladocs/action-linkspector@37c85bcde51b30bf929936502bac6bfb7e8f0a4d # v1.4.1

AGENTS.md

@@ -110,6 +110,9 @@ app, err := api.Database.GetOAuth2ProviderAppByClientID(ctx, clientID)
 - For experimental or unstable API paths, skip public doc generation with
   `// @x-apidocgen {"skip": true}` after the `@Router` annotation. This
   keeps them out of the published API reference until they stabilize.
+- Experimental chat endpoints in `coderd/exp_chats.go` omit swagger
+  annotations entirely. Do not add `@Summary`, `@Router`, or other
+  swagger comments to handlers in that file.
 
 ### Database Query Naming
 

Makefile

@@ -988,6 +988,7 @@ coderd/httpmw/loggermw/loggermock/loggermock.go: coderd/httpmw/loggermw/logger.g
 
 codersdk/workspacesdk/agentconnmock/agentconnmock.go: codersdk/workspacesdk/agentconn.go
 	go generate ./codersdk/workspacesdk/agentconnmock/
+	./scripts/format_go_file.sh "$@"
 	touch "$@"
 
 $(AIBRIDGED_MOCKS): enterprise/aibridged/client.go enterprise/aibridged/pool.go

agent/agent.go

@@ -38,6 +38,7 @@ import (
 	"cdr.dev/slog/v3"
 	"github.com/coder/clistat"
 	"github.com/coder/coder/v2/agent/agentcontainers"
+	"github.com/coder/coder/v2/agent/agentcontextconfig"
 	"github.com/coder/coder/v2/agent/agentexec"
 	"github.com/coder/coder/v2/agent/agentfiles"
 	"github.com/coder/coder/v2/agent/agentgit"
@@ -308,12 +309,13 @@ type agent struct {
 	containerAPI        *agentcontainers.API
 	gitAPIOptions       []agentgit.Option
 
-	filesAPI   *agentfiles.API
-	gitAPI     *agentgit.API
-	processAPI *agentproc.API
-	desktopAPI *agentdesktop.API
-	mcpManager *agentmcp.Manager
-	mcpAPI     *agentmcp.API
+	filesAPI         *agentfiles.API
+	gitAPI           *agentgit.API
+	processAPI       *agentproc.API
+	desktopAPI       *agentdesktop.API
+	mcpManager       *agentmcp.Manager
+	mcpAPI           *agentmcp.API
+	contextConfigAPI *agentcontextconfig.API
 
 	socketServerEnabled bool
 	socketPath          string
@@ -396,11 +398,17 @@ func (a *agent) init() {
 	gitOpts := append([]agentgit.Option{agentgit.WithClock(a.clock)}, a.gitAPIOptions...)
 	a.gitAPI = agentgit.NewAPI(a.logger.Named("git"), pathStore, gitOpts...)
 	desktop := agentdesktop.NewPortableDesktop(
-		a.logger.Named("desktop"), a.execer, a.scriptRunner.ScriptBinDir(),
+		a.logger.Named("desktop"), a.execer, a.scriptRunner.ScriptBinDir(), nil,
 	)
 	a.desktopAPI = agentdesktop.NewAPI(a.logger.Named("desktop"), desktop, a.clock)
 	a.mcpManager = agentmcp.NewManager(a.logger.Named("mcp"))
 	a.mcpAPI = agentmcp.NewAPI(a.logger.Named("mcp"), a.mcpManager)
+	a.contextConfigAPI = agentcontextconfig.NewAPI(func() string {
+		if m := a.manifest.Load(); m != nil {
+			return m.Directory
+		}
+		return ""
+	})
 	a.reconnectingPTYServer = reconnectingpty.NewServer(
 		a.logger.Named("reconnecting-pty"),
 		a.sshServer,
@@ -1358,7 +1366,7 @@ func (a *agent) handleManifest(manifestOK *checkpoint) func(ctx context.Context,
 				// lifecycle transition to avoid delaying Ready.
 				// This runs inside the tracked goroutine so it
 				// is properly awaited on shutdown.
-				if mcpErr := a.mcpManager.Connect(a.gracefulCtx, manifest.Directory); mcpErr != nil {
+				if mcpErr := a.mcpManager.Connect(a.gracefulCtx, a.contextConfigAPI.MCPConfigFiles()); mcpErr != nil {
 					a.logger.Warn(ctx, "failed to connect to workspace MCP servers", slog.Error(mcpErr))
 				}
 			})

agent/agent_internal_test.go

@@ -1,6 +1,8 @@
 package agent
 
 import (
+	"path/filepath"
+	"runtime"
 	"testing"
 
 	"github.com/google/uuid"
@@ -8,10 +10,22 @@ import (
 
 	"cdr.dev/slog/v3"
 	"cdr.dev/slog/v3/sloggers/slogtest"
+	"github.com/coder/coder/v2/agent/agentcontextconfig"
 	"github.com/coder/coder/v2/agent/proto"
+	agentsdk "github.com/coder/coder/v2/codersdk/agentsdk"
 	"github.com/coder/coder/v2/testutil"
 )
 
+// platformAbsPath constructs an absolute path that is valid
+// on the current platform. On Windows, paths must include a
+// drive letter to be considered absolute.
+func platformAbsPath(parts ...string) string {
+	if runtime.GOOS == "windows" {
+		return `C:\` + filepath.Join(parts...)
+	}
+	return "/" + filepath.Join(parts...)
+}
+
 // 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) {
@@ -42,3 +56,41 @@ func TestReportConnectionEmpty(t *testing.T) {
 	require.Equal(t, proto.Connection_DISCONNECT, req1.GetConnection().GetAction())
 	require.Equal(t, "because", req1.GetConnection().GetReason())
 }
+
+func TestContextConfigAPI_InitOnce(t *testing.T) {
+	// Not parallel: uses t.Setenv to clear env vars.
+
+	// Clear env vars so defaults are used and the test is
+	// hermetic regardless of the surrounding environment.
+	t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+	t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+	t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+	t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+	t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+	// After the fix, contextConfigAPI is set once in init() and
+	// never reassigned. Config() evaluates lazily via the
+	// manifest, so there is no concurrent write to race with.
+	dir1 := platformAbsPath("dir1")
+	dir2 := platformAbsPath("dir2")
+
+	a := &agent{}
+	a.manifest.Store(&agentsdk.Manifest{Directory: dir1})
+	a.contextConfigAPI = agentcontextconfig.NewAPI(func() string {
+		if m := a.manifest.Load(); m != nil {
+			return m.Directory
+		}
+		return ""
+	})
+
+	mcpFiles1 := a.contextConfigAPI.MCPConfigFiles()
+	require.NotEmpty(t, mcpFiles1)
+	require.Contains(t, mcpFiles1[0], dir1)
+
+	// Simulate manifest update on reconnection -- no field
+	// reassignment needed, the lazy closure picks it up.
+	a.manifest.Store(&agentsdk.Manifest{Directory: dir2})
+	mcpFiles2 := a.contextConfigAPI.MCPConfigFiles()
+	require.NotEmpty(t, mcpFiles2)
+	require.Contains(t, mcpFiles2[0], dir2)
+}

agent/agent_test.go

@@ -3007,7 +3007,7 @@ func TestAgent_Speedtest(t *testing.T) {
 
 func TestAgent_Reconnect(t *testing.T) {
 	t.Parallel()
-	ctx := testutil.Context(t, testutil.WaitShort)
+	ctx := testutil.Context(t, testutil.WaitLong)
 	logger := testutil.Logger(t)
 	// After the agent is disconnected from a coordinator, it's supposed
 	// to reconnect!
@@ -3020,7 +3020,8 @@ func TestAgent_Reconnect(t *testing.T) {
 		logger,
 		agentID,
 		agentsdk.Manifest{
-			DERPMap: derpMap,
+			DERPMap:   derpMap,
+			Directory: "/test/workspace",
 		},
 		statsCh,
 		fCoordinator,
@@ -3033,13 +3034,19 @@ func TestAgent_Reconnect(t *testing.T) {
 	})
 	defer closer.Close()
 
-	call1 := testutil.RequireReceive(ctx, t, fCoordinator.CoordinateCalls)
-	require.Equal(t, client.GetNumRefreshTokenCalls(), 1)
-	close(call1.Resps) // hang up
-	// expect reconnect
+	// Each iteration forces the agent to reconnect by closing
+	// the current coordinate call while the tracked HTTP server
+	// goroutine (from connection 1's createTailnet) is still
+	// alive, widening the race window.
+	const reconnections = 5
+	for i := range reconnections {
+		call := testutil.RequireReceive(ctx, t, fCoordinator.CoordinateCalls)
+		require.Equal(t, i+1, client.GetNumRefreshTokenCalls())
+		close(call.Resps) // hang up — triggers reconnect
+	}
+	// Verify final reconnect succeeds.
 	testutil.RequireReceive(ctx, t, fCoordinator.CoordinateCalls)
-	// Check that the agent refreshes the token when it reconnects.
-	require.Equal(t, client.GetNumRefreshTokenCalls(), 2)
+	require.Equal(t, reconnections+1, client.GetNumRefreshTokenCalls())
 	closer.Close()
 }
 

agent/agentcontextconfig/api.go

@@ -0,0 +1,313 @@
+package agentcontextconfig
+
+import (
+	"cmp"
+	"io"
+	"net/http"
+	"os"
+	"path/filepath"
+	"regexp"
+	"strings"
+
+	"github.com/go-chi/chi/v5"
+
+	"github.com/coder/coder/v2/coderd/httpapi"
+	"github.com/coder/coder/v2/codersdk"
+	"github.com/coder/coder/v2/codersdk/workspacesdk"
+)
+
+// Env var names for context configuration. Prefixed with EXP_
+// to indicate these are experimental and may change.
+const (
+	EnvInstructionsDirs = "CODER_AGENT_EXP_INSTRUCTIONS_DIRS"
+	EnvInstructionsFile = "CODER_AGENT_EXP_INSTRUCTIONS_FILE"
+	EnvSkillsDirs       = "CODER_AGENT_EXP_SKILLS_DIRS"
+	EnvSkillMetaFile    = "CODER_AGENT_EXP_SKILL_META_FILE"
+	EnvMCPConfigFiles   = "CODER_AGENT_EXP_MCP_CONFIG_FILES"
+)
+
+const (
+	maxInstructionFileBytes = 64 * 1024
+	maxSkillMetaBytes       = 64 * 1024
+)
+
+// markdownCommentPattern strips HTML comments from instruction
+// file content for security (prevents hidden prompt injection).
+var markdownCommentPattern = regexp.MustCompile(`<!--[\s\S]*?-->`)
+
+// invisibleRunePattern strips invisible Unicode characters that
+// could be used for prompt injection.
+//
+//nolint:gocritic // Non-ASCII char ranges are intentional for invisible Unicode stripping.
+var invisibleRunePattern = regexp.MustCompile(
+	"[\u00ad\u034f\u061c\u070f" +
+		"\u115f\u1160\u17b4\u17b5" +
+		"\u180b-\u180f" +
+		"\u200b\u200d\u200e\u200f" +
+		"\u202a-\u202e" +
+		"\u2060-\u206f" +
+		"\u3164" +
+		"\ufe00-\ufe0f" +
+		"\ufeff" +
+		"\uffa0" +
+		"\ufff0-\ufff8]",
+)
+
+// skillNamePattern validates kebab-case skill names.
+var skillNamePattern = regexp.MustCompile(
+	`^[a-z0-9]+(-[a-z0-9]+)*$`,
+)
+
+// Default values for agent-internal configuration. These are
+// used when the corresponding env vars are unset.
+const (
+	DefaultInstructionsDir  = "~/.coder"
+	DefaultInstructionsFile = "AGENTS.md"
+	DefaultSkillsDir        = ".agents/skills"
+	DefaultSkillMetaFile    = "SKILL.md"
+	DefaultMCPConfigFile    = ".mcp.json"
+)
+
+// API exposes the resolved context configuration through the
+// agent's HTTP API.
+type API struct {
+	workingDir func() string
+}
+
+// NewAPI accepts a closure that returns the working directory.
+// The directory is evaluated lazily on each call to Config(),
+// so the caller can update it after construction.
+func NewAPI(workingDir func() string) *API {
+	if workingDir == nil {
+		workingDir = func() string { return "" }
+	}
+	return &API{workingDir: workingDir}
+}
+
+// Config reads env vars, resolves paths, reads instruction files,
+// and discovers skills. Returns the HTTP response and the resolved
+// MCP config file paths (used only agent-internally). Exported
+// for use by tests.
+func Config(workingDir string) (workspacesdk.ContextConfigResponse, []string) {
+	// TrimSpace all env vars before cmp.Or so that a
+	// whitespace-only value falls through to the default
+	// consistently. ResolvePaths also trims each comma-
+	// separated entry, but without pre-trimming here a
+	// bare " " would bypass cmp.Or and produce nil.
+	instructionsDir := cmp.Or(strings.TrimSpace(os.Getenv(EnvInstructionsDirs)), DefaultInstructionsDir)
+	instructionsFile := cmp.Or(strings.TrimSpace(os.Getenv(EnvInstructionsFile)), DefaultInstructionsFile)
+	skillsDir := cmp.Or(strings.TrimSpace(os.Getenv(EnvSkillsDirs)), DefaultSkillsDir)
+	skillMetaFile := cmp.Or(strings.TrimSpace(os.Getenv(EnvSkillMetaFile)), DefaultSkillMetaFile)
+	mcpConfigFile := cmp.Or(strings.TrimSpace(os.Getenv(EnvMCPConfigFiles)), DefaultMCPConfigFile)
+
+	resolvedInstructionsDirs := ResolvePaths(instructionsDir, workingDir)
+	resolvedSkillsDirs := ResolvePaths(skillsDir, workingDir)
+
+	// Read instruction files from each configured directory.
+	parts := readInstructionFiles(resolvedInstructionsDirs, instructionsFile)
+
+	// Also check the working directory for the instruction file,
+	// unless it was already covered by InstructionsDirs.
+	if workingDir != "" {
+		seenDirs := make(map[string]struct{}, len(resolvedInstructionsDirs))
+		for _, d := range resolvedInstructionsDirs {
+			seenDirs[d] = struct{}{}
+		}
+		if _, ok := seenDirs[workingDir]; !ok {
+			if entry, found := readInstructionFileFromDir(workingDir, instructionsFile); found {
+				parts = append(parts, entry)
+			}
+		}
+	}
+
+	// Discover skills from each configured skills directory.
+	skillParts := discoverSkills(resolvedSkillsDirs, skillMetaFile)
+	parts = append(parts, skillParts...)
+
+	// Guarantee non-nil slice to signal agent support.
+	if parts == nil {
+		parts = []codersdk.ChatMessagePart{}
+	}
+
+	return workspacesdk.ContextConfigResponse{
+		Parts: parts,
+	}, ResolvePaths(mcpConfigFile, workingDir)
+}
+
+// MCPConfigFiles returns the resolved MCP configuration file
+// paths for the agent's MCP manager.
+func (api *API) MCPConfigFiles() []string {
+	_, mcpFiles := Config(api.workingDir())
+	return mcpFiles
+}
+
+// Routes returns the HTTP handler for the context config
+// endpoint.
+func (api *API) Routes() http.Handler {
+	r := chi.NewRouter()
+	r.Get("/", api.handleGet)
+	return r
+}
+
+func (api *API) handleGet(rw http.ResponseWriter, r *http.Request) {
+	response, _ := Config(api.workingDir())
+	httpapi.Write(r.Context(), rw, http.StatusOK, response)
+}
+
+// readInstructionFiles reads instruction files from each given
+// directory. Missing directories are silently skipped. Duplicate
+// directories are deduplicated.
+func readInstructionFiles(dirs []string, fileName string) []codersdk.ChatMessagePart {
+	var parts []codersdk.ChatMessagePart
+	seen := make(map[string]struct{}, len(dirs))
+	for _, dir := range dirs {
+		if _, ok := seen[dir]; ok {
+			continue
+		}
+		seen[dir] = struct{}{}
+		if part, found := readInstructionFileFromDir(dir, fileName); found {
+			parts = append(parts, part)
+		}
+	}
+	return parts
+}
+
+// readInstructionFileFromDir scans a directory for a file matching
+// fileName (case-insensitive) and reads its contents.
+func readInstructionFileFromDir(dir, fileName string) (codersdk.ChatMessagePart, bool) {
+	dirEntries, err := os.ReadDir(dir)
+	if err != nil {
+		return codersdk.ChatMessagePart{}, false
+	}
+
+	for _, e := range dirEntries {
+		if e.IsDir() {
+			continue
+		}
+		if strings.EqualFold(strings.TrimSpace(e.Name()), fileName) {
+			filePath := filepath.Join(dir, e.Name())
+			content, truncated, ok := readAndSanitizeFile(filePath, maxInstructionFileBytes)
+			if !ok {
+				return codersdk.ChatMessagePart{}, false
+			}
+			if content == "" {
+				return codersdk.ChatMessagePart{}, false
+			}
+			return codersdk.ChatMessagePart{
+				Type:                 codersdk.ChatMessagePartTypeContextFile,
+				ContextFilePath:      filePath,
+				ContextFileContent:   content,
+				ContextFileTruncated: truncated,
+			}, true
+		}
+	}
+	return codersdk.ChatMessagePart{}, false
+}
+
+// readAndSanitizeFile reads the file at path, capping the read
+// at maxBytes to avoid unbounded memory allocation. It sanitizes
+// the content (strips HTML comments and invisible Unicode) and
+// returns the result. Returns false if the file cannot be read.
+func readAndSanitizeFile(path string, maxBytes int64) (content string, truncated bool, ok bool) {
+	f, err := os.Open(path)
+	if err != nil {
+		return "", false, false
+	}
+	defer f.Close()
+
+	// Read at most maxBytes+1 to detect truncation without
+	// allocating the entire file into memory.
+	raw, err := io.ReadAll(io.LimitReader(f, maxBytes+1))
+	if err != nil {
+		return "", false, false
+	}
+
+	truncated = int64(len(raw)) > maxBytes
+	if truncated {
+		raw = raw[:maxBytes]
+	}
+
+	s := sanitizeInstructionMarkdown(string(raw))
+	if s == "" {
+		return "", truncated, true
+	}
+	return s, truncated, true
+}
+
+// sanitizeInstructionMarkdown strips HTML comments, invisible
+// Unicode characters, and CRLF line endings from instruction
+// file content.
+func sanitizeInstructionMarkdown(content string) string {
+	content = strings.ReplaceAll(content, "\r\n", "\n")
+	content = strings.ReplaceAll(content, "\r", "\n")
+	content = markdownCommentPattern.ReplaceAllString(content, "")
+	content = invisibleRunePattern.ReplaceAllString(content, "")
+	return strings.TrimSpace(content)
+}
+
+// discoverSkills walks the given skills directories and returns
+// metadata for every valid skill it finds. Body and supporting
+// file lists are NOT included; chatd fetches those on demand
+// via read_skill. Missing directories or individual errors are
+// silently skipped.
+func discoverSkills(skillsDirs []string, metaFile string) []codersdk.ChatMessagePart {
+	seen := make(map[string]struct{})
+	var parts []codersdk.ChatMessagePart
+
+	for _, skillsDir := range skillsDirs {
+		entries, err := os.ReadDir(skillsDir)
+		if err != nil {
+			continue
+		}
+
+		for _, entry := range entries {
+			if !entry.IsDir() {
+				continue
+			}
+
+			metaPath := filepath.Join(skillsDir, entry.Name(), metaFile)
+			f, err := os.Open(metaPath)
+			if err != nil {
+				continue
+			}
+			raw, err := io.ReadAll(io.LimitReader(f, maxSkillMetaBytes+1))
+			_ = f.Close()
+			if err != nil {
+				continue
+			}
+			if int64(len(raw)) > maxSkillMetaBytes {
+				raw = raw[:maxSkillMetaBytes]
+			}
+
+			name, description, _, err := workspacesdk.ParseSkillFrontmatter(string(raw))
+			if err != nil {
+				continue
+			}
+
+			// The directory name must match the declared name.
+			if name != entry.Name() {
+				continue
+			}
+			if !skillNamePattern.MatchString(name) {
+				continue
+			}
+
+			// First occurrence wins across directories.
+			if _, ok := seen[name]; ok {
+				continue
+			}
+			seen[name] = struct{}{}
+
+			skillDir := filepath.Join(skillsDir, entry.Name())
+			parts = append(parts, codersdk.ChatMessagePart{
+				Type:                     codersdk.ChatMessagePartTypeSkill,
+				SkillName:                name,
+				SkillDescription:         description,
+				SkillDir:                 skillDir,
+				ContextFileSkillMetaFile: metaFile,
+			})
+		}
+	}
+
+	return parts
+}

agent/agentcontextconfig/api_test.go

@@ -0,0 +1,439 @@
+package agentcontextconfig_test
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+
+	"github.com/coder/coder/v2/agent/agentcontextconfig"
+	"github.com/coder/coder/v2/codersdk"
+)
+
+// filterParts returns only the parts matching the given type.
+func filterParts(parts []codersdk.ChatMessagePart, t codersdk.ChatMessagePartType) []codersdk.ChatMessagePart {
+	var out []codersdk.ChatMessagePart
+	for _, p := range parts {
+		if p.Type == t {
+			out = append(out, p)
+		}
+	}
+	return out
+}
+
+func TestConfig(t *testing.T) {
+	t.Run("Defaults", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+
+		// Clear all env vars so defaults are used.
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := platformAbsPath("work")
+		cfg, mcpFiles := agentcontextconfig.Config(workDir)
+
+		// Parts is always non-nil.
+		require.NotNil(t, cfg.Parts)
+		// Default MCP config file is ".mcp.json" (relative),
+		// resolved against the working directory.
+		require.Equal(t, []string{filepath.Join(workDir, ".mcp.json")}, mcpFiles)
+	})
+
+	t.Run("CustomEnvVars", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+
+		optInstructions := t.TempDir()
+		optSkills := t.TempDir()
+		optMCP := platformAbsPath("opt", "mcp.json")
+
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, optInstructions)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "CUSTOM.md")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, optSkills)
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "META.yaml")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, optMCP)
+
+		// Create files matching the custom names so we can
+		// verify the env vars actually change lookup behavior.
+		require.NoError(t, os.WriteFile(filepath.Join(optInstructions, "CUSTOM.md"), []byte("custom instructions"), 0o600))
+		skillDir := filepath.Join(optSkills, "my-skill")
+		require.NoError(t, os.MkdirAll(skillDir, 0o755))
+		require.NoError(t, os.WriteFile(
+			filepath.Join(skillDir, "META.yaml"),
+			[]byte("---\nname: my-skill\ndescription: custom meta\n---\n"),
+			0o600,
+		))
+
+		workDir := platformAbsPath("work")
+		cfg, mcpFiles := agentcontextconfig.Config(workDir)
+
+		require.Equal(t, []string{optMCP}, mcpFiles)
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.Len(t, ctxFiles, 1)
+		require.Equal(t, "custom instructions", ctxFiles[0].ContextFileContent)
+		skillParts := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeSkill)
+		require.Len(t, skillParts, 1)
+		require.Equal(t, "my-skill", skillParts[0].SkillName)
+		require.Equal(t, "META.yaml", skillParts[0].ContextFileSkillMetaFile)
+	})
+
+	t.Run("WhitespaceInFileNames", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "  CLAUDE.md  ")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		// Create a file matching the trimmed name.
+		require.NoError(t, os.WriteFile(filepath.Join(fakeHome, "CLAUDE.md"), []byte("hello"), 0o600))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.Len(t, ctxFiles, 1)
+		require.Equal(t, "hello", ctxFiles[0].ContextFileContent)
+	})
+
+	t.Run("CommaSeparatedDirs", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+
+		a := t.TempDir()
+		b := t.TempDir()
+
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, a+","+b)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		// Put instruction files in both dirs.
+		require.NoError(t, os.WriteFile(filepath.Join(a, "AGENTS.md"), []byte("from a"), 0o600))
+		require.NoError(t, os.WriteFile(filepath.Join(b, "AGENTS.md"), []byte("from b"), 0o600))
+
+		workDir := t.TempDir()
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.Len(t, ctxFiles, 2)
+		require.Equal(t, "from a", ctxFiles[0].ContextFileContent)
+		require.Equal(t, "from b", ctxFiles[1].ContextFileContent)
+	})
+
+	t.Run("ReadsInstructionFiles", func(t *testing.T) {
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+
+		// Create ~/.coder/AGENTS.md
+		coderDir := filepath.Join(fakeHome, ".coder")
+		require.NoError(t, os.MkdirAll(coderDir, 0o755))
+		require.NoError(t, os.WriteFile(
+			filepath.Join(coderDir, "AGENTS.md"),
+			[]byte("home instructions"),
+			0o600,
+		))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.NotNil(t, cfg.Parts)
+		require.Len(t, ctxFiles, 1)
+		require.Equal(t, "home instructions", ctxFiles[0].ContextFileContent)
+		require.Equal(t, filepath.Join(coderDir, "AGENTS.md"), ctxFiles[0].ContextFilePath)
+		require.False(t, ctxFiles[0].ContextFileTruncated)
+	})
+
+	t.Run("ReadsWorkingDirInstructionFile", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+
+		// Create AGENTS.md in the working directory.
+		require.NoError(t, os.WriteFile(
+			filepath.Join(workDir, "AGENTS.md"),
+			[]byte("project instructions"),
+			0o600,
+		))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		// Should find the working dir file (not in instruction dirs).
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.NotNil(t, cfg.Parts)
+		require.Len(t, ctxFiles, 1)
+		require.Equal(t, "project instructions", ctxFiles[0].ContextFileContent)
+		require.Equal(t, filepath.Join(workDir, "AGENTS.md"), ctxFiles[0].ContextFilePath)
+	})
+
+	t.Run("TruncatesLargeInstructionFile", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		largeContent := strings.Repeat("a", 64*1024+100)
+		require.NoError(t, os.WriteFile(filepath.Join(workDir, "AGENTS.md"), []byte(largeContent), 0o600))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.Len(t, ctxFiles, 1)
+		require.True(t, ctxFiles[0].ContextFileTruncated)
+		require.Len(t, ctxFiles[0].ContextFileContent, 64*1024)
+	})
+
+	t.Run("SanitizesHTMLComments", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		require.NoError(t, os.WriteFile(
+			filepath.Join(workDir, "AGENTS.md"),
+			[]byte("visible\n<!-- hidden -->content"),
+			0o600,
+		))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.Len(t, ctxFiles, 1)
+		require.Equal(t, "visible\ncontent", ctxFiles[0].ContextFileContent)
+	})
+
+	t.Run("SanitizesInvisibleUnicode", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		// U+200B (zero-width space) should be stripped.
+		require.NoError(t, os.WriteFile(
+			filepath.Join(workDir, "AGENTS.md"),
+			[]byte("before\u200bafter"),
+			0o600,
+		))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.Len(t, ctxFiles, 1)
+		require.Equal(t, "beforeafter", ctxFiles[0].ContextFileContent)
+	})
+
+	t.Run("NormalizesCRLF", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		require.NoError(t, os.WriteFile(
+			filepath.Join(workDir, "AGENTS.md"),
+			[]byte("line1\r\nline2\rline3"),
+			0o600,
+		))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		ctxFiles := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeContextFile)
+		require.Len(t, ctxFiles, 1)
+		require.Equal(t, "line1\nline2\nline3", ctxFiles[0].ContextFileContent)
+	})
+
+	t.Run("DiscoversSkills", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		skillsDir := filepath.Join(workDir, ".agents", "skills")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, skillsDir)
+
+		// Create a valid skill.
+		skillDir := filepath.Join(skillsDir, "my-skill")
+		require.NoError(t, os.MkdirAll(skillDir, 0o755))
+		require.NoError(t, os.WriteFile(
+			filepath.Join(skillDir, "SKILL.md"),
+			[]byte("---\nname: my-skill\ndescription: A test skill\n---\nSkill body"),
+			0o600,
+		))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		skillParts := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeSkill)
+		require.Len(t, skillParts, 1)
+		require.Equal(t, "my-skill", skillParts[0].SkillName)
+		require.Equal(t, "A test skill", skillParts[0].SkillDescription)
+		require.Equal(t, skillDir, skillParts[0].SkillDir)
+		require.Equal(t, "SKILL.md", skillParts[0].ContextFileSkillMetaFile)
+	})
+
+	t.Run("SkipsMissingDirs", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+
+		nonExistent := filepath.Join(t.TempDir(), "does-not-exist")
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, nonExistent)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, nonExistent)
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		cfg, _ := agentcontextconfig.Config(workDir)
+
+		// Non-nil empty slice (signals agent supports new format).
+		require.NotNil(t, cfg.Parts)
+		require.Empty(t, cfg.Parts)
+	})
+
+	t.Run("MCPConfigFilesResolvedSeparately", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+
+		optMCP := platformAbsPath("opt", "custom.json")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, optMCP)
+
+		workDir := t.TempDir()
+		_, mcpFiles := agentcontextconfig.Config(workDir)
+
+		require.Equal(t, []string{optMCP}, mcpFiles)
+	})
+
+	t.Run("SkillNameMustMatchDir", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		skillsDir := filepath.Join(workDir, "skills")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, skillsDir)
+
+		// Skill name in frontmatter doesn't match directory name.
+		skillDir := filepath.Join(skillsDir, "wrong-dir-name")
+		require.NoError(t, os.MkdirAll(skillDir, 0o755))
+		require.NoError(t, os.WriteFile(
+			filepath.Join(skillDir, "SKILL.md"),
+			[]byte("---\nname: actual-name\ndescription: mismatch\n---\n"),
+			0o600,
+		))
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+		skillParts := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeSkill)
+		require.Empty(t, skillParts)
+	})
+
+	t.Run("DuplicateSkillsFirstWins", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsDirs, fakeHome)
+		t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+		t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+		t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+		workDir := t.TempDir()
+		skillsDir1 := filepath.Join(workDir, "skills1")
+		skillsDir2 := filepath.Join(workDir, "skills2")
+		t.Setenv(agentcontextconfig.EnvSkillsDirs, skillsDir1+","+skillsDir2)
+
+		// Same skill name in both directories.
+		for _, dir := range []string{skillsDir1, skillsDir2} {
+			skillDir := filepath.Join(dir, "dup-skill")
+			require.NoError(t, os.MkdirAll(skillDir, 0o755))
+			require.NoError(t, os.WriteFile(
+				filepath.Join(skillDir, "SKILL.md"),
+				[]byte("---\nname: dup-skill\ndescription: from "+filepath.Base(dir)+"\n---\n"),
+				0o600,
+			))
+		}
+
+		cfg, _ := agentcontextconfig.Config(workDir)
+		skillParts := filterParts(cfg.Parts, codersdk.ChatMessagePartTypeSkill)
+		require.Len(t, skillParts, 1)
+		require.Equal(t, "from skills1", skillParts[0].SkillDescription)
+	})
+}
+
+func TestNewAPI_LazyDirectory(t *testing.T) {
+	t.Setenv(agentcontextconfig.EnvInstructionsDirs, "")
+	t.Setenv(agentcontextconfig.EnvInstructionsFile, "")
+	t.Setenv(agentcontextconfig.EnvSkillsDirs, "")
+	t.Setenv(agentcontextconfig.EnvSkillMetaFile, "")
+	t.Setenv(agentcontextconfig.EnvMCPConfigFiles, "")
+
+	dir := ""
+	api := agentcontextconfig.NewAPI(func() string { return dir })
+
+	// Before directory is set, MCP paths resolve to nothing.
+	mcpFiles := api.MCPConfigFiles()
+	require.Empty(t, mcpFiles)
+
+	// After setting the directory, MCPConfigFiles() picks it up.
+	dir = platformAbsPath("work")
+	mcpFiles = api.MCPConfigFiles()
+	require.NotEmpty(t, mcpFiles)
+	require.Equal(t, []string{filepath.Join(dir, ".mcp.json")}, mcpFiles)
+}

agent/agentcontextconfig/resolve.go

@@ -0,0 +1,55 @@
+package agentcontextconfig
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+)
+
+// ResolvePath resolves a single path that may be absolute,
+// home-relative (~/ or ~), or relative to the given base
+// directory. Returns an absolute path. Empty input returns empty.
+func ResolvePath(raw, baseDir string) string {
+	raw = strings.TrimSpace(raw)
+	if raw == "" {
+		return ""
+	}
+	switch {
+	case raw == "~":
+		home, err := os.UserHomeDir()
+		if err != nil {
+			return ""
+		}
+		return home
+	case strings.HasPrefix(raw, "~/"):
+		home, err := os.UserHomeDir()
+		if err != nil {
+			return ""
+		}
+		return filepath.Join(home, raw[2:])
+	case filepath.IsAbs(raw):
+		return raw
+	default:
+		if baseDir == "" {
+			return ""
+		}
+		return filepath.Join(baseDir, raw)
+	}
+}
+
+// ResolvePaths splits a comma-separated list of paths and
+// resolves each entry independently. Empty entries and entries
+// that resolve to empty strings are skipped.
+func ResolvePaths(raw, baseDir string) []string {
+	if strings.TrimSpace(raw) == "" {
+		return nil
+	}
+	parts := strings.Split(raw, ",")
+	out := make([]string, 0, len(parts))
+	for _, p := range parts {
+		if resolved := ResolvePath(p, baseDir); resolved != "" {
+			out = append(out, resolved)
+		}
+	}
+	return out
+}

agent/agentcontextconfig/resolve_test.go

@@ -0,0 +1,152 @@
+package agentcontextconfig_test
+
+import (
+	"path/filepath"
+	"runtime"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+
+	"github.com/coder/coder/v2/agent/agentcontextconfig"
+)
+
+// platformAbsPath constructs an absolute path that is valid
+// on the current platform. On Windows paths must include a
+// drive letter to be considered absolute.
+func platformAbsPath(parts ...string) string {
+	if runtime.GOOS == "windows" {
+		return `C:\` + filepath.Join(parts...)
+	}
+	return "/" + filepath.Join(parts...)
+}
+
+func TestResolvePath(t *testing.T) { //nolint:tparallel // subtests using t.Setenv cannot be parallel
+	t.Run("EmptyInput", func(t *testing.T) {
+		t.Parallel()
+		require.Equal(t, "", agentcontextconfig.ResolvePath("", platformAbsPath("base")))
+	})
+
+	t.Run("WhitespaceOnly", func(t *testing.T) {
+		t.Parallel()
+		require.Equal(t, "", agentcontextconfig.ResolvePath("   ", platformAbsPath("base")))
+	})
+
+	// Tests that use t.Setenv cannot be parallel.
+	t.Run("TildeAlone", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		got := agentcontextconfig.ResolvePath("~", platformAbsPath("base"))
+		require.Equal(t, fakeHome, got)
+	})
+
+	t.Run("TildeSlashPath", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		got := agentcontextconfig.ResolvePath("~/docs/readme", platformAbsPath("base"))
+		require.Equal(t, filepath.Join(fakeHome, "docs", "readme"), got)
+	})
+
+	t.Run("AbsolutePath", func(t *testing.T) {
+		t.Parallel()
+		p := platformAbsPath("etc", "coder")
+		got := agentcontextconfig.ResolvePath(p, platformAbsPath("base"))
+		require.Equal(t, p, got)
+	})
+
+	t.Run("RelativePath", func(t *testing.T) {
+		t.Parallel()
+		base := platformAbsPath("work")
+		got := agentcontextconfig.ResolvePath("foo/bar", base)
+		require.Equal(t, filepath.Join(base, "foo", "bar"), got)
+	})
+
+	t.Run("RelativePathWithWhitespace", func(t *testing.T) {
+		t.Parallel()
+		base := platformAbsPath("work")
+		got := agentcontextconfig.ResolvePath("  foo/bar  ", base)
+		require.Equal(t, filepath.Join(base, "foo", "bar"), got)
+	})
+
+	t.Run("RelativePathWithEmptyBaseDir", func(t *testing.T) {
+		t.Parallel()
+		got := agentcontextconfig.ResolvePath(".agents/skills", "")
+		require.Equal(t, "", got)
+	})
+}
+
+func TestResolvePath_HomeUnset(t *testing.T) {
+	// Cannot be parallel — modifies HOME env var.
+	t.Setenv("HOME", "")
+	// Also clear USERPROFILE for Windows compatibility.
+	t.Setenv("USERPROFILE", "")
+
+	require.Equal(t, "", agentcontextconfig.ResolvePath("~", platformAbsPath("base")))
+	require.Equal(t, "", agentcontextconfig.ResolvePath("~/docs", platformAbsPath("base")))
+}
+
+func TestResolvePaths(t *testing.T) { //nolint:tparallel // subtests using t.Setenv cannot be parallel
+	t.Run("EmptyString", func(t *testing.T) {
+		t.Parallel()
+		require.Nil(t, agentcontextconfig.ResolvePaths("", platformAbsPath("base")))
+	})
+
+	t.Run("WhitespaceOnly", func(t *testing.T) {
+		t.Parallel()
+		require.Nil(t, agentcontextconfig.ResolvePaths("   ", platformAbsPath("base")))
+	})
+
+	t.Run("SingleEntry", func(t *testing.T) {
+		t.Parallel()
+		p := platformAbsPath("abs", "path")
+		got := agentcontextconfig.ResolvePaths(p, platformAbsPath("base"))
+		require.Equal(t, []string{p}, got)
+	})
+
+	// Tests that use t.Setenv cannot be parallel.
+	t.Run("MultipleEntries", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		b := platformAbsPath("b")
+		base := platformAbsPath("base")
+		got := agentcontextconfig.ResolvePaths("~/a,"+b+",rel", base)
+		require.Equal(t, []string{
+			filepath.Join(fakeHome, "a"),
+			b,
+			filepath.Join(base, "rel"),
+		}, got)
+	})
+
+	t.Run("TrimsWhitespace", func(t *testing.T) {
+		t.Parallel()
+		a := platformAbsPath("a")
+		b := platformAbsPath("b")
+		got := agentcontextconfig.ResolvePaths("  "+a+" , "+b+"  ", platformAbsPath("base"))
+		require.Equal(t, []string{a, b}, got)
+	})
+
+	t.Run("SkipsEmptyEntries", func(t *testing.T) {
+		t.Parallel()
+		a := platformAbsPath("a")
+		b := platformAbsPath("b")
+		got := agentcontextconfig.ResolvePaths(a+",,"+b+",", platformAbsPath("base"))
+		require.Equal(t, []string{a, b}, got)
+	})
+
+	t.Run("TrailingComma", func(t *testing.T) {
+		t.Parallel()
+		p := platformAbsPath("only")
+		got := agentcontextconfig.ResolvePaths(p+",", platformAbsPath("base"))
+		require.Equal(t, []string{p}, got)
+	})
+
+	t.Run("RelativePathSkippedWhenBaseDirEmpty", func(t *testing.T) {
+		fakeHome := t.TempDir()
+		t.Setenv("HOME", fakeHome)
+		t.Setenv("USERPROFILE", fakeHome)
+		got := agentcontextconfig.ResolvePaths("~/.coder,.agents/skills", "")
+		require.Equal(t, []string{filepath.Join(fakeHome, ".coder")}, got)
+	})
+}

agent/agentproc/process.go

@@ -148,6 +148,11 @@ func (m *manager) start(req workspacesdk.StartProcessRequest, chatID string) (*p
 	for k, v := range req.Env {
 		cmd.Env = append(cmd.Env, fmt.Sprintf("%s=%s", k, v))
 	}
+	// Propagate the chat ID so child processes (e.g.
+	// GIT_ASKPASS) can send it back to the server.
+	if chatID != "" {
+		cmd.Env = append(cmd.Env, fmt.Sprintf("CODER_CHAT_ID=%s", chatID))
+	}
 
 	if err := cmd.Start(); err != nil {
 		cancel()

agent/agentssh/x11_test.go

@@ -211,7 +211,7 @@ func TestServer_X11_EvictionLRU(t *testing.T) {
 		require.NoError(t, err)
 		stderr, err := sess.StderrPipe()
 		require.NoError(t, err)
-		require.NoError(t, sess.Shell())
+		require.NoError(t, sess.Start("sh"))
 
 		// The SSH server lazily starts the session. We need to write a command
 		// and read back to ensure the X11 forwarding is started.

agent/api.go

@@ -32,6 +32,7 @@ func (a *agent) apiHandler() http.Handler {
 	r.Mount("/api/v0/processes", a.processAPI.Routes())
 	r.Mount("/api/v0/desktop", a.desktopAPI.Routes())
 	r.Mount("/api/v0/mcp", a.mcpAPI.Routes())
+	r.Mount("/api/v0/context-config", a.contextConfigAPI.Routes())
 
 	if a.devcontainers {
 		r.Mount("/api/v0/containers", a.containerAPI.Routes())

agent/x/agentdesktop/api.go

@@ -1,12 +1,17 @@
 package agentdesktop
 
 import (
+	"context"
 	"encoding/json"
+	"errors"
+	"io"
 	"net/http"
 	"strconv"
+	"sync"
 	"time"
 
 	"github.com/go-chi/chi/v5"
+	"github.com/google/uuid"
 
 	"cdr.dev/slog/v3"
 	"github.com/coder/coder/v2/agent/agentssh"
@@ -47,6 +52,9 @@ type API struct {
 	logger  slog.Logger
 	desktop Desktop
 	clock   quartz.Clock
+
+	closeMu sync.Mutex
+	closed  bool
 }
 
 // NewAPI creates a new desktop streaming API.
@@ -66,6 +74,10 @@ func (a *API) Routes() http.Handler {
 	r := chi.NewRouter()
 	r.Get("/vnc", a.handleDesktopVNC)
 	r.Post("/action", a.handleAction)
+	r.Route("/recording", func(r chi.Router) {
+		r.Post("/start", a.handleRecordingStart)
+		r.Post("/stop", a.handleRecordingStop)
+	})
 	return r
 }
 
@@ -116,6 +128,9 @@ func (a *API) handleAction(rw http.ResponseWriter, r *http.Request) {
 	ctx := r.Context()
 	handlerStart := a.clock.Now()
 
+	// Update last desktop action timestamp for idle recording monitor.
+	a.desktop.RecordActivity()
+
 	// Ensure the desktop is running and grab native dimensions.
 	cfg, err := a.desktop.Start(ctx)
 	if err != nil {
@@ -480,9 +495,150 @@ func (a *API) handleAction(rw http.ResponseWriter, r *http.Request) {
 
 // Close shuts down the desktop session if one is running.
 func (a *API) Close() error {
+	a.closeMu.Lock()
+	if a.closed {
+		a.closeMu.Unlock()
+		return nil
+	}
+	a.closed = true
+	a.closeMu.Unlock()
+
 	return a.desktop.Close()
 }
 
+// decodeRecordingRequest decodes and validates a recording request
+// from the HTTP body, returning the recording ID. Returns false if
+// the request was invalid and an error response was already written.
+func (*API) decodeRecordingRequest(rw http.ResponseWriter, r *http.Request) (string, bool) {
+	ctx := r.Context()
+	var req struct {
+		RecordingID string `json:"recording_id"`
+	}
+	if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
+		httpapi.Write(ctx, rw, http.StatusBadRequest, codersdk.Response{
+			Message: "Failed to decode request body.",
+			Detail:  err.Error(),
+		})
+		return "", false
+	}
+	if req.RecordingID == "" {
+		httpapi.Write(ctx, rw, http.StatusBadRequest, codersdk.Response{
+			Message: "Missing recording_id.",
+		})
+		return "", false
+	}
+	if _, err := uuid.Parse(req.RecordingID); err != nil {
+		httpapi.Write(ctx, rw, http.StatusBadRequest, codersdk.Response{
+			Message: "Invalid recording_id format.",
+			Detail:  "recording_id must be a valid UUID.",
+		})
+		return "", false
+	}
+	return req.RecordingID, true
+}
+
+func (a *API) handleRecordingStart(rw http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+	recordingID, ok := a.decodeRecordingRequest(rw, r)
+	if !ok {
+		return
+	}
+
+	a.closeMu.Lock()
+	if a.closed {
+		a.closeMu.Unlock()
+		httpapi.Write(ctx, rw, http.StatusServiceUnavailable, codersdk.Response{
+			Message: "Desktop API is shutting down.",
+		})
+		return
+	}
+	a.closeMu.Unlock()
+
+	if err := a.desktop.StartRecording(ctx, recordingID); err != nil {
+		if errors.Is(err, ErrDesktopClosed) {
+			httpapi.Write(ctx, rw, http.StatusServiceUnavailable, codersdk.Response{
+				Message: "Desktop API is shutting down.",
+			})
+			return
+		}
+		httpapi.Write(ctx, rw, http.StatusInternalServerError, codersdk.Response{
+			Message: "Failed to start recording.",
+			Detail:  err.Error(),
+		})
+		return
+	}
+
+	httpapi.Write(ctx, rw, http.StatusOK, codersdk.Response{
+		Message: "Recording started.",
+	})
+}
+
+func (a *API) handleRecordingStop(rw http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	recordingID, ok := a.decodeRecordingRequest(rw, r)
+	if !ok {
+		return
+	}
+
+	a.closeMu.Lock()
+	if a.closed {
+		a.closeMu.Unlock()
+		httpapi.Write(ctx, rw, http.StatusServiceUnavailable, codersdk.Response{
+			Message: "Desktop API is shutting down.",
+		})
+		return
+	}
+	a.closeMu.Unlock()
+
+	// Stop recording (idempotent).
+	// Use a context detached from the HTTP request so that if the
+	// connection drops, the recording process can still shut down
+	// gracefully. WithoutCancel preserves request-scoped values.
+	stopCtx, stopCancel := context.WithTimeout(context.WithoutCancel(r.Context()), 30*time.Second)
+	defer stopCancel()
+	artifact, err := a.desktop.StopRecording(stopCtx, recordingID)
+	if err != nil {
+		if errors.Is(err, ErrUnknownRecording) {
+			httpapi.Write(ctx, rw, http.StatusNotFound, codersdk.Response{
+				Message: "Recording not found.",
+				Detail:  err.Error(),
+			})
+			return
+		}
+		if errors.Is(err, ErrRecordingCorrupted) {
+			httpapi.Write(ctx, rw, http.StatusInternalServerError, codersdk.Response{
+				Message: "Recording is corrupted.",
+				Detail:  err.Error(),
+			})
+			return
+		}
+		httpapi.Write(ctx, rw, http.StatusInternalServerError, codersdk.Response{
+			Message: "Failed to stop recording.",
+			Detail:  err.Error(),
+		})
+		return
+	}
+	defer artifact.Reader.Close()
+
+	if artifact.Size > workspacesdk.MaxRecordingSize {
+		a.logger.Warn(ctx, "recording file exceeds maximum size",
+			slog.F("recording_id", recordingID),
+			slog.F("size", artifact.Size),
+			slog.F("max_size", workspacesdk.MaxRecordingSize),
+		)
+		httpapi.Write(ctx, rw, http.StatusRequestEntityTooLarge, codersdk.Response{
+			Message: "Recording file exceeds maximum allowed size.",
+		})
+		return
+	}
+
+	rw.Header().Set("Content-Type", "video/mp4")
+	rw.Header().Set("Content-Length", strconv.FormatInt(artifact.Size, 10))
+	rw.WriteHeader(http.StatusOK)
+	_, _ = io.Copy(rw, artifact.Reader)
+}
+
 // coordFromAction extracts the coordinate pair from a DesktopAction,
 // returning an error if the coordinate field is missing.
 func coordFromAction(action DesktopAction) (x, y int, err error) {

agent/x/agentdesktop/api_test.go

@@ -4,12 +4,17 @@ import (
 	"bytes"
 	"context"
 	"encoding/json"
+	"fmt"
 	"net"
 	"net/http"
 	"net/http/httptest"
+	"os"
+	"slices"
+	"sync"
 	"testing"
 	"time"
 
+	"github.com/google/uuid"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 	"golang.org/x/xerrors"
@@ -21,6 +26,16 @@ import (
 	"github.com/coder/quartz"
 )
 
+// Test recording UUIDs used across tests.
+const (
+	testRecIDDefault         = "870e1f02-8118-4300-a37e-4adb0117baf3"
+	testRecIDStartIdempotent = "250a2ffb-a5e5-4c94-9754-4d6a4ab7ba20"
+	testRecIDStopIdempotent  = "38f8a378-f98f-4758-a4ae-950b44cf989a"
+	testRecIDConcurrentA     = "8dc173eb-23c6-4601-a485-b6dfb2a42c3a"
+	testRecIDConcurrentB     = "fea490d4-70f0-4798-a181-29d65ce25ae1"
+	testRecIDRestart         = "75173a0d-b018-4e2e-a771-defa3fc6af69"
+)
+
 // Ensure fakeDesktop satisfies the Desktop interface at compile time.
 var _ agentdesktop.Desktop = (*fakeDesktop)(nil)
 
@@ -43,6 +58,14 @@ type fakeDesktop struct {
 	lastTyped   string
 	lastKeyDown string
 	lastKeyUp   string
+
+	// Recording tracking (guarded by recMu).
+	recMu         sync.Mutex
+	recordings    map[string]string // ID → file path
+	stopCalls     []string          // recording IDs passed to StopRecording
+	recStopCh     chan string       // optional: signaled when StopRecording is called
+	startCount    int               // incremented on each new recording start
+	activityCount int               // incremented by RecordActivity
 }
 
 func (f *fakeDesktop) Start(context.Context) (agentdesktop.DisplayConfig, error) {
@@ -107,11 +130,140 @@ func (f *fakeDesktop) CursorPosition(context.Context) (x int, y int, err error)
 	return f.cursorPos[0], f.cursorPos[1], nil
 }
 
+func (f *fakeDesktop) StartRecording(_ context.Context, recordingID string) error {
+	f.recMu.Lock()
+	defer f.recMu.Unlock()
+	if f.recordings == nil {
+		f.recordings = make(map[string]string)
+	}
+	if path, ok := f.recordings[recordingID]; ok {
+		// Check if already stopped (file still exists but stop was
+		// called). For the fake, a stopped recording means its ID
+		// appears in stopCalls. In that case, remove the old file
+		// and start fresh.
+		stopped := slices.Contains(f.stopCalls, recordingID)
+		if !stopped {
+			// Active recording - no-op.
+			return nil
+		}
+		// Completed recording - discard old file, start fresh.
+		_ = os.Remove(path)
+		delete(f.recordings, recordingID)
+	}
+	f.startCount++
+	tmpFile, err := os.CreateTemp("", "fake-recording-*.mp4")
+	if err != nil {
+		return err
+	}
+	_, _ = tmpFile.Write([]byte(fmt.Sprintf("fake-mp4-data-%s-%d", recordingID, f.startCount)))
+	_ = tmpFile.Close()
+	f.recordings[recordingID] = tmpFile.Name()
+	return nil
+}
+
+func (f *fakeDesktop) StopRecording(_ context.Context, recordingID string) (*agentdesktop.RecordingArtifact, error) {
+	f.recMu.Lock()
+	defer f.recMu.Unlock()
+	if f.recordings == nil {
+		return nil, agentdesktop.ErrUnknownRecording
+	}
+	path, ok := f.recordings[recordingID]
+	if !ok {
+		return nil, agentdesktop.ErrUnknownRecording
+	}
+	f.stopCalls = append(f.stopCalls, recordingID)
+	if f.recStopCh != nil {
+		select {
+		case f.recStopCh <- recordingID:
+		default:
+		}
+	}
+	file, err := os.Open(path)
+	if err != nil {
+		return nil, err
+	}
+	info, err := file.Stat()
+	if err != nil {
+		_ = file.Close()
+		return nil, err
+	}
+	return &agentdesktop.RecordingArtifact{
+		Reader: file,
+		Size:   info.Size(),
+	}, nil
+}
+
+func (f *fakeDesktop) RecordActivity() {
+	f.recMu.Lock()
+	f.activityCount++
+	f.recMu.Unlock()
+}
+
 func (f *fakeDesktop) Close() error {
 	f.closed = true
+	f.recMu.Lock()
+	defer f.recMu.Unlock()
+	for _, path := range f.recordings {
+		_ = os.Remove(path)
+	}
 	return nil
 }
 
+// failStartRecordingDesktop wraps fakeDesktop and overrides
+// StartRecording to always return an error.
+type failStartRecordingDesktop struct {
+	fakeDesktop
+	startRecordingErr error
+}
+
+func (f *failStartRecordingDesktop) StartRecording(_ context.Context, _ string) error {
+	return f.startRecordingErr
+}
+
+// corruptedStopDesktop wraps fakeDesktop and overrides
+// StopRecording to always return ErrRecordingCorrupted.
+type corruptedStopDesktop struct {
+	fakeDesktop
+}
+
+func (*corruptedStopDesktop) StopRecording(_ context.Context, _ string) (*agentdesktop.RecordingArtifact, error) {
+	return nil, agentdesktop.ErrRecordingCorrupted
+}
+
+// oversizedFakeDesktop wraps fakeDesktop and expands recording files
+// beyond MaxRecordingSize when StopRecording is called.
+type oversizedFakeDesktop struct {
+	fakeDesktop
+}
+
+func (f *oversizedFakeDesktop) StopRecording(ctx context.Context, recordingID string) (*agentdesktop.RecordingArtifact, error) {
+	artifact, err := f.fakeDesktop.StopRecording(ctx, recordingID)
+	if err != nil {
+		return nil, err
+	}
+	// Close the original reader since we're going to re-open after truncation.
+	artifact.Reader.Close()
+
+	// Look up the path from the fakeDesktop recordings.
+	f.fakeDesktop.recMu.Lock()
+	path := f.fakeDesktop.recordings[recordingID]
+	f.fakeDesktop.recMu.Unlock()
+
+	// Expand the file to exceed the maximum recording size.
+	if err := os.Truncate(path, workspacesdk.MaxRecordingSize+1); err != nil {
+		return nil, err
+	}
+	// Re-open the truncated file.
+	file, err := os.Open(path)
+	if err != nil {
+		return nil, err
+	}
+	return &agentdesktop.RecordingArtifact{
+		Reader: file,
+		Size:   workspacesdk.MaxRecordingSize + 1,
+	}, nil
+}
+
 func TestHandleDesktopVNC_StartError(t *testing.T) {
 	t.Parallel()
 
@@ -134,6 +286,37 @@ func TestHandleDesktopVNC_StartError(t *testing.T) {
 	assert.Equal(t, "Failed to start desktop session.", resp.Message)
 }
 
+func TestHandleAction_CallsRecordActivity(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	body := agentdesktop.DesktopAction{
+		Action:     "left_click",
+		Coordinate: &[2]int{100, 200},
+	}
+	b, err := json.Marshal(body)
+	require.NoError(t, err)
+
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/action", bytes.NewReader(b))
+	req.Header.Set("Content-Type", "application/json")
+
+	handler := api.Routes()
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+
+	fake.recMu.Lock()
+	count := fake.activityCount
+	fake.recMu.Unlock()
+	assert.Equal(t, 1, count, "handleAction should call RecordActivity exactly once")
+}
+
 func TestHandleAction_Screenshot(t *testing.T) {
 	t.Parallel()
 
@@ -574,3 +757,481 @@ func TestHandleAction_CursorPositionReturnsDeclaredCoordinates(t *testing.T) {
 	// Native (960,540) in 1920x1080 should map to declared space in 1280x720.
 	assert.Equal(t, "x=640,y=360", resp.Output)
 }
+
+func TestRecordingStartStop(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Start recording.
+	startBody, err := json.Marshal(map[string]string{"recording_id": testRecIDDefault})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(startBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+
+	// Stop recording.
+	stopBody, err := json.Marshal(map[string]string{"recording_id": testRecIDDefault})
+	require.NoError(t, err)
+	rr = httptest.NewRecorder()
+	req = httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(stopBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+	assert.Equal(t, "video/mp4", rr.Header().Get("Content-Type"))
+	assert.Equal(t, []byte("fake-mp4-data-"+testRecIDDefault+"-1"), rr.Body.Bytes())
+}
+
+func TestRecordingStartFails(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &failStartRecordingDesktop{
+		fakeDesktop: fakeDesktop{
+			startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+		},
+		startRecordingErr: xerrors.New("start recording error"),
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	body, err := json.Marshal(map[string]string{"recording_id": uuid.New().String()})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(body))
+	handler.ServeHTTP(rr, req)
+
+	assert.Equal(t, http.StatusInternalServerError, rr.Code)
+
+	var resp codersdk.Response
+	err = json.NewDecoder(rr.Body).Decode(&resp)
+	require.NoError(t, err)
+	assert.Equal(t, "Failed to start recording.", resp.Message)
+}
+
+func TestRecordingStartIdempotent(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Start same recording twice - both should succeed.
+	for range 2 {
+		body, err := json.Marshal(map[string]string{"recording_id": testRecIDStartIdempotent})
+		require.NoError(t, err)
+		rr := httptest.NewRecorder()
+		req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(body))
+		handler.ServeHTTP(rr, req)
+		require.Equal(t, http.StatusOK, rr.Code)
+	}
+
+	// Stop once, verify normal response.
+	stopBody, err := json.Marshal(map[string]string{"recording_id": testRecIDStartIdempotent})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(stopBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+	assert.Equal(t, "video/mp4", rr.Header().Get("Content-Type"))
+	assert.Equal(t, []byte("fake-mp4-data-"+testRecIDStartIdempotent+"-1"), rr.Body.Bytes())
+}
+
+func TestRecordingStopIdempotent(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Start recording.
+	startBody, err := json.Marshal(map[string]string{"recording_id": testRecIDStopIdempotent})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(startBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+
+	// Stop twice - both should succeed with identical data.
+	var bodies [2][]byte
+	for i := range 2 {
+		body, err := json.Marshal(map[string]string{"recording_id": testRecIDStopIdempotent})
+		require.NoError(t, err)
+		recorder := httptest.NewRecorder()
+		request := httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(body))
+		handler.ServeHTTP(recorder, request)
+		require.Equal(t, http.StatusOK, recorder.Code)
+		assert.Equal(t, "video/mp4", recorder.Header().Get("Content-Type"))
+		bodies[i] = recorder.Body.Bytes()
+	}
+	assert.Equal(t, bodies[0], bodies[1])
+}
+
+func TestRecordingStopInvalidIDFormat(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	body, err := json.Marshal(map[string]string{"recording_id": "not-a-uuid"})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(body))
+	handler.ServeHTTP(rr, req)
+	assert.Equal(t, http.StatusBadRequest, rr.Code)
+}
+
+func TestRecordingStopUnknownRecording(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Send a valid UUID that was never started - should reach
+	// StopRecording, get ErrUnknownRecording, and return 404.
+	body, err := json.Marshal(map[string]string{"recording_id": uuid.New().String()})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(body))
+	handler.ServeHTTP(rr, req)
+	assert.Equal(t, http.StatusNotFound, rr.Code)
+
+	var resp codersdk.Response
+	err = json.NewDecoder(rr.Body).Decode(&resp)
+	require.NoError(t, err)
+	assert.Equal(t, "Recording not found.", resp.Message)
+}
+
+func TestRecordingStopOversizedFile(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &oversizedFakeDesktop{
+		fakeDesktop: fakeDesktop{
+			startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+		},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Start recording.
+	recID := uuid.New().String()
+	startBody, err := json.Marshal(map[string]string{"recording_id": recID})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(startBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+
+	// Stop recording - file exceeds max size, expect 413.
+	stopBody, err := json.Marshal(map[string]string{"recording_id": recID})
+	require.NoError(t, err)
+	rr = httptest.NewRecorder()
+	req = httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(stopBody))
+	handler.ServeHTTP(rr, req)
+	assert.Equal(t, http.StatusRequestEntityTooLarge, rr.Code)
+
+	var resp codersdk.Response
+	err = json.NewDecoder(rr.Body).Decode(&resp)
+	require.NoError(t, err)
+	assert.Equal(t, "Recording file exceeds maximum allowed size.", resp.Message)
+}
+
+func TestRecordingMultipleSimultaneous(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Start two recordings with different IDs.
+	for _, id := range []string{testRecIDConcurrentA, testRecIDConcurrentB} {
+		body, err := json.Marshal(map[string]string{"recording_id": id})
+		require.NoError(t, err)
+		rr := httptest.NewRecorder()
+		req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(body))
+		handler.ServeHTTP(rr, req)
+		require.Equal(t, http.StatusOK, rr.Code)
+	}
+
+	// Stop both and verify each returns its own data.
+	expected := map[string][]byte{
+		testRecIDConcurrentA: []byte("fake-mp4-data-" + testRecIDConcurrentA + "-1"),
+		testRecIDConcurrentB: []byte("fake-mp4-data-" + testRecIDConcurrentB + "-2"),
+	}
+	for _, id := range []string{testRecIDConcurrentA, testRecIDConcurrentB} {
+		body, err := json.Marshal(map[string]string{"recording_id": id})
+		require.NoError(t, err)
+		rr := httptest.NewRecorder()
+		req := httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(body))
+		handler.ServeHTTP(rr, req)
+		require.Equal(t, http.StatusOK, rr.Code)
+		assert.Equal(t, "video/mp4", rr.Header().Get("Content-Type"))
+		assert.Equal(t, expected[id], rr.Body.Bytes())
+	}
+}
+
+func TestRecordingStartMalformedBody(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader([]byte("not json")))
+	handler.ServeHTTP(rr, req)
+	assert.Equal(t, http.StatusBadRequest, rr.Code)
+}
+
+func TestRecordingStartEmptyID(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	body, err := json.Marshal(map[string]string{"recording_id": ""})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(body))
+	handler.ServeHTTP(rr, req)
+	assert.Equal(t, http.StatusBadRequest, rr.Code)
+}
+
+func TestRecordingStopEmptyID(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	body, err := json.Marshal(map[string]string{"recording_id": ""})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(body))
+	handler.ServeHTTP(rr, req)
+	assert.Equal(t, http.StatusBadRequest, rr.Code)
+}
+
+func TestRecordingStopMalformedBody(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader([]byte("not json")))
+	handler.ServeHTTP(rr, req)
+	assert.Equal(t, http.StatusBadRequest, rr.Code)
+}
+
+func TestRecordingStartAfterCompleted(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Step 1: Start recording.
+	startBody, err := json.Marshal(map[string]string{"recording_id": testRecIDRestart})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(startBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+
+	// Step 2: Stop recording (gets first MP4 data).
+	stopBody, err := json.Marshal(map[string]string{"recording_id": testRecIDRestart})
+	require.NoError(t, err)
+	rr = httptest.NewRecorder()
+	req = httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(stopBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+	assert.Equal(t, "video/mp4", rr.Header().Get("Content-Type"))
+	firstData := rr.Body.Bytes()
+	require.NotEmpty(t, firstData)
+
+	// Step 3: Start again with the same ID - should succeed
+	// (old file discarded, new recording started).
+	rr = httptest.NewRecorder()
+	req = httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(startBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+
+	// Step 4: Stop again - should return NEW MP4 data.
+	rr = httptest.NewRecorder()
+	req = httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(stopBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+	assert.Equal(t, "video/mp4", rr.Header().Get("Content-Type"))
+	secondData := rr.Body.Bytes()
+	require.NotEmpty(t, secondData)
+
+	// The two recordings should have different data because the
+	// fake increments a counter on each fresh start.
+	assert.NotEqual(t, firstData, secondData,
+		"restarted recording should produce different data")
+}
+
+func TestRecordingStartAfterClose(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &fakeDesktop{
+		startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+
+	handler := api.Routes()
+
+	// Close the API before sending the request.
+	api.Close()
+
+	body, err := json.Marshal(map[string]string{"recording_id": uuid.New().String()})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(body))
+	handler.ServeHTTP(rr, req)
+
+	assert.Equal(t, http.StatusServiceUnavailable, rr.Code)
+
+	var resp codersdk.Response
+	err = json.NewDecoder(rr.Body).Decode(&resp)
+	require.NoError(t, err)
+	assert.Equal(t, "Desktop API is shutting down.", resp.Message)
+}
+
+func TestRecordingStartDesktopClosed(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	// StartRecording returns ErrDesktopClosed to simulate a race
+	// where the desktop is closed between the API-level check and
+	// the desktop-level StartRecording call.
+	fake := &failStartRecordingDesktop{
+		fakeDesktop: fakeDesktop{
+			startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+		},
+		startRecordingErr: agentdesktop.ErrDesktopClosed,
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	body, err := json.Marshal(map[string]string{"recording_id": uuid.New().String()})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(body))
+	handler.ServeHTTP(rr, req)
+
+	assert.Equal(t, http.StatusServiceUnavailable, rr.Code)
+
+	var resp codersdk.Response
+	err = json.NewDecoder(rr.Body).Decode(&resp)
+	require.NoError(t, err)
+	assert.Equal(t, "Desktop API is shutting down.", resp.Message)
+}
+
+func TestRecordingStopCorrupted(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	fake := &corruptedStopDesktop{
+		fakeDesktop: fakeDesktop{
+			startCfg: agentdesktop.DisplayConfig{Width: 1920, Height: 1080},
+		},
+	}
+	api := agentdesktop.NewAPI(logger, fake, nil)
+	defer api.Close()
+
+	handler := api.Routes()
+
+	// Start a recording so the stop has something to find.
+	recID := uuid.New().String()
+	startBody, err := json.Marshal(map[string]string{"recording_id": recID})
+	require.NoError(t, err)
+	rr := httptest.NewRecorder()
+	req := httptest.NewRequest(http.MethodPost, "/recording/start", bytes.NewReader(startBody))
+	handler.ServeHTTP(rr, req)
+	require.Equal(t, http.StatusOK, rr.Code)
+
+	// Stop returns ErrRecordingCorrupted.
+	stopBody, err := json.Marshal(map[string]string{"recording_id": recID})
+	require.NoError(t, err)
+	rr = httptest.NewRecorder()
+	req = httptest.NewRequest(http.MethodPost, "/recording/stop", bytes.NewReader(stopBody))
+	handler.ServeHTTP(rr, req)
+
+	assert.Equal(t, http.StatusInternalServerError, rr.Code)
+
+	var respStop codersdk.Response
+	err = json.NewDecoder(rr.Body).Decode(&respStop)
+	require.NoError(t, err)
+	assert.Equal(t, "Recording is corrupted.", respStop.Message)
+}

agent/x/agentdesktop/desktop.go

@@ -2,7 +2,10 @@ package agentdesktop
 
 import (
 	"context"
+	"io"
 	"net"
+
+	"golang.org/x/xerrors"
 )
 
 // Desktop abstracts a virtual desktop session running inside a workspace.
@@ -58,10 +61,52 @@ type Desktop interface {
 	// CursorPosition returns the current cursor coordinates.
 	CursorPosition(ctx context.Context) (x, y int, err error)
 
+	// RecordActivity marks the desktop as having received user
+	// interaction, resetting the idle-recording timer.
+	RecordActivity()
+
+	// StartRecording begins recording the desktop to an MP4 file
+	// using the caller-provided recording ID. Safe to call
+	// repeatedly - active recordings continue unchanged, stopped
+	// recordings are discarded and restarted. Concurrent recordings
+	// are supported.
+	StartRecording(ctx context.Context, recordingID string) error
+
+	// StopRecording finalizes the recording identified by the given
+	// ID. Idempotent - safe to call on an already-stopped recording.
+	// Returns a RecordingArtifact that the caller can stream. The
+	// caller must close the artifact when done. Returns an error if
+	// the recording ID is unknown.
+	StopRecording(ctx context.Context, recordingID string) (*RecordingArtifact, error)
+
 	// Close shuts down the desktop session and cleans up resources.
 	Close() error
 }
 
+// ErrUnknownRecording is returned by StopRecording when the
+// recording ID is not recognized.
+var ErrUnknownRecording = xerrors.New("unknown recording ID")
+
+// ErrDesktopClosed is returned when an operation is attempted on a
+// closed desktop session.
+var ErrDesktopClosed = xerrors.New("desktop closed")
+
+// ErrRecordingCorrupted is returned by StopRecording when the
+// recording process was force-killed and the artifact is likely
+// incomplete or corrupt.
+var ErrRecordingCorrupted = xerrors.New("recording corrupted: process was force-killed")
+
+// RecordingArtifact is a finalized recording returned by StopRecording.
+// The caller streams the artifact and must call Close when done. The
+// artifact remains valid even if the same recording ID is restarted
+// or the desktop is closed while the caller is reading.
+type RecordingArtifact struct {
+	// Reader is the MP4 content. Callers must close it when done.
+	Reader io.ReadCloser
+	// Size is the byte length of the MP4 content.
+	Size int64
+}
+
 // DisplayConfig describes a running desktop session.
 type DisplayConfig struct {
 	Width   int // native width in pixels

agent/x/agentdesktop/portabledesktop.go

@@ -3,6 +3,7 @@ package agentdesktop
 import (
 	"context"
 	"encoding/json"
+	"errors"
 	"fmt"
 	"net"
 	"os"
@@ -11,6 +12,7 @@ import (
 	"runtime"
 	"strconv"
 	"sync"
+	"sync/atomic"
 	"time"
 
 	"golang.org/x/xerrors"
@@ -18,6 +20,7 @@ import (
 	"cdr.dev/slog/v3"
 	"github.com/coder/coder/v2/agent/agentexec"
 	"github.com/coder/coder/v2/codersdk/workspacesdk"
+	"github.com/coder/quartz"
 )
 
 // portableDesktopOutput is the JSON output from
@@ -49,32 +52,65 @@ type screenshotOutput struct {
 	Data string `json:"data"`
 }
 
+// recordingProcess tracks a single desktop recording subprocess.
+type recordingProcess struct {
+	cmd        *exec.Cmd
+	filePath   string
+	stopped    bool
+	killed     bool          // true when the process was SIGKILLed
+	done       chan struct{} // closed when cmd.Wait() returns
+	waitErr    error         // set before done is closed
+	stopOnce   sync.Once
+	idleCancel context.CancelFunc // cancels the per-recording idle goroutine
+	idleDone   chan struct{}      // closed when idle goroutine exits
+}
+
+// maxConcurrentRecordings is the maximum number of active (non-stopped)
+// recordings allowed at once. This prevents resource exhaustion.
+const maxConcurrentRecordings = 5
+
+// idleTimeout is the duration of desktop inactivity after which all
+// active recordings are automatically stopped.
+const idleTimeout = 10 * time.Minute
+
 // portableDesktop implements Desktop by shelling out to the
 // portabledesktop CLI via agentexec.Execer.
 type portableDesktop struct {
 	logger       slog.Logger
 	execer       agentexec.Execer
 	scriptBinDir string // coder script bin directory
+	clock        quartz.Clock
 
-	mu      sync.Mutex
-	session *desktopSession // nil until started
-	binPath string          // resolved path to binary, cached
-	closed  bool
+	mu                  sync.Mutex
+	session             *desktopSession // nil until started
+	binPath             string          // resolved path to binary, cached
+	closed              bool
+	recordings          map[string]*recordingProcess // guarded by mu
+	lastDesktopActionAt atomic.Int64
 }
 
 // NewPortableDesktop creates a Desktop backed by the portabledesktop
 // CLI binary, using execer to spawn child processes. scriptBinDir is
-// the coder script bin directory checked for the binary.
+// the coder script bin directory checked for the binary. If clk is
+// nil, a real clock is used.
 func NewPortableDesktop(
 	logger slog.Logger,
 	execer agentexec.Execer,
 	scriptBinDir string,
+	clk quartz.Clock,
 ) Desktop {
-	return &portableDesktop{
+	if clk == nil {
+		clk = quartz.NewReal()
+	}
+	pd := &portableDesktop{
 		logger:       logger,
 		execer:       execer,
 		scriptBinDir: scriptBinDir,
+		clock:        clk,
+		recordings:   make(map[string]*recordingProcess),
 	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+	return pd
 }
 
 // Start launches the desktop session (idempotent).
@@ -83,7 +119,7 @@ func (p *portableDesktop) Start(ctx context.Context) (DisplayConfig, error) {
 	defer p.mu.Unlock()
 
 	if p.closed {
-		return DisplayConfig{}, xerrors.New("desktop is closed")
+		return DisplayConfig{}, ErrDesktopClosed
 	}
 
 	if err := p.ensureBinary(ctx); err != nil {
@@ -313,23 +349,328 @@ func (p *portableDesktop) CursorPosition(ctx context.Context) (x int, y int, err
 	return result.X, result.Y, nil
 }
 
-// Close shuts down the desktop session and cleans up resources.
-func (p *portableDesktop) Close() error {
+// StartRecording begins recording the desktop to an MP4 file.
+// Three-state idempotency: active recordings are no-ops,
+// completed recordings are discarded and restarted.
+func (p *portableDesktop) StartRecording(ctx context.Context, recordingID string) error {
+	// Ensure the desktop session is running before acquiring the
+	// recording lock. Start is independently locked and idempotent.
+	if _, err := p.Start(ctx); err != nil {
+		return xerrors.Errorf("ensure desktop session: %w", err)
+	}
+
 	p.mu.Lock()
 	defer p.mu.Unlock()
 
+	if p.closed {
+		return ErrDesktopClosed
+	}
+
+	// Three-state idempotency:
+	// - Active recording → no-op, continue recording.
+	// - Completed recording → discard old file, start fresh.
+	// - Unknown ID → fall through to start a new recording.
+	if rec, ok := p.recordings[recordingID]; ok {
+		if !rec.stopped {
+			select {
+			case <-rec.done:
+				// Process exited unexpectedly; treat as completed
+				// so we fall through to discard the old file and
+				// restart.
+			default:
+				// Active recording - no-op, continue recording.
+				return nil
+			}
+		}
+		// Completed recording - discard old file, start fresh.
+		if err := os.Remove(rec.filePath); err != nil && !os.IsNotExist(err) {
+			p.logger.Warn(ctx, "failed to remove old recording file",
+				slog.F("recording_id", recordingID),
+				slog.F("file_path", rec.filePath),
+				slog.Error(err),
+			)
+		}
+		delete(p.recordings, recordingID)
+	}
+
+	// Check concurrent recording limit.
+	if p.lockedActiveRecordingCount() >= maxConcurrentRecordings {
+		return xerrors.Errorf("too many concurrent recordings (max %d)", maxConcurrentRecordings)
+	}
+
+	// GC sweep: remove stopped recordings with stale files.
+	p.lockedCleanStaleRecordings(ctx)
+
+	if err := p.ensureBinary(ctx); err != nil {
+		return xerrors.Errorf("ensure portabledesktop binary: %w", err)
+	}
+
+	filePath := filepath.Join(os.TempDir(), "coder-recording-"+recordingID+".mp4")
+
+	// Use a background context so the process outlives the HTTP
+	// request that triggered it.
+	procCtx, procCancel := context.WithCancel(context.Background())
+
+	//nolint:gosec // portabledesktop is a trusted binary resolved via ensureBinary.
+	cmd := p.execer.CommandContext(procCtx, p.binPath, "record",
+		// The following options are used to speed up the recording when the desktop is idle.
+		// They were taken out of an example in the portabledesktop repo.
+		// There's likely room for improvement to optimize the values.
+		"--idle-speedup", "20",
+		"--idle-min-duration", "0.35",
+		"--idle-noise-tolerance", "-38dB",
+		filePath)
+
+	if err := cmd.Start(); err != nil {
+		procCancel()
+		return xerrors.Errorf("start recording process: %w", err)
+	}
+
+	rec := &recordingProcess{
+		cmd:      cmd,
+		filePath: filePath,
+		done:     make(chan struct{}),
+	}
+	go func() {
+		rec.waitErr = cmd.Wait()
+		close(rec.done)
+		// avoid a context resource leak by canceling the context
+		procCancel()
+	}()
+
+	p.recordings[recordingID] = rec
+
+	p.logger.Info(ctx, "started desktop recording",
+		slog.F("recording_id", recordingID),
+		slog.F("file_path", filePath),
+		slog.F("pid", cmd.Process.Pid),
+	)
+
+	// Record activity so a recording started on an already-idle
+	// desktop does not stop immediately.
+	p.lastDesktopActionAt.Store(p.clock.Now().UnixNano())
+
+	// Spawn a per-recording idle goroutine.
+	idleCtx, idleCancel := context.WithCancel(context.Background())
+	rec.idleCancel = idleCancel
+	rec.idleDone = make(chan struct{})
+	go func() {
+		defer close(rec.idleDone)
+		p.monitorRecordingIdle(idleCtx, rec)
+	}()
+
+	return nil
+}
+
+// StopRecording finalizes the recording. Idempotent - safe to call
+// on an already-stopped recording. Returns a RecordingArtifact
+// that the caller can stream. The caller must close the Reader
+// on the returned artifact to avoid leaking file descriptors.
+func (p *portableDesktop) StopRecording(ctx context.Context, recordingID string) (*RecordingArtifact, error) {
+	p.mu.Lock()
+	rec, ok := p.recordings[recordingID]
+	if !ok {
+		p.mu.Unlock()
+		return nil, ErrUnknownRecording
+	}
+
+	p.lockedStopRecordingProcess(ctx, rec, false)
+	killed := rec.killed
+	p.mu.Unlock()
+
+	p.logger.Info(ctx, "stopped desktop recording",
+		slog.F("recording_id", recordingID),
+		slog.F("file_path", rec.filePath),
+	)
+
+	if killed {
+		return nil, ErrRecordingCorrupted
+	}
+
+	// Open the file and return an artifact. Each call opens a fresh
+	// file descriptor so the caller is insulated from restarts and
+	// desktop close.
+	f, err := os.Open(rec.filePath)
+	if err != nil {
+		return nil, xerrors.Errorf("open recording artifact: %w", err)
+	}
+	info, err := f.Stat()
+	if err != nil {
+		_ = f.Close()
+		return nil, xerrors.Errorf("stat recording artifact: %w", err)
+	}
+	return &RecordingArtifact{
+		Reader: f,
+		Size:   info.Size(),
+	}, nil
+}
+
+// lockedStopRecordingProcess stops a single recording via stopOnce.
+// It sends SIGINT, waits up to 15 seconds for graceful exit, then
+// SIGKILLs. When force is true the process is SIGKILLed immediately
+// without attempting a graceful shutdown. Must be called while p.mu
+// is held; the lock is held for the full duration so that no
+// concurrent StopRecording caller can read rec.stopped = true
+// before the process has finished writing the MP4 file.
+//
+//nolint:revive // force flag keeps shared stopOnce/cleanup logic in one place.
+func (p *portableDesktop) lockedStopRecordingProcess(ctx context.Context, rec *recordingProcess, force bool) {
+	rec.stopOnce.Do(func() {
+		if force {
+			_ = rec.cmd.Process.Kill()
+			rec.killed = true
+		} else {
+			_ = interruptRecordingProcess(rec.cmd.Process)
+			timer := p.clock.NewTimer(15*time.Second, "agentdesktop", "stop_timeout")
+			defer timer.Stop()
+			select {
+			case <-rec.done:
+			case <-ctx.Done():
+				_ = rec.cmd.Process.Kill()
+				rec.killed = true
+			case <-timer.C:
+				_ = rec.cmd.Process.Kill()
+				rec.killed = true
+			}
+		}
+		rec.stopped = true
+		if rec.idleCancel != nil {
+			rec.idleCancel()
+		}
+	})
+	// NOTE: We intentionally do not wait on rec.done here.
+	// If goleak is added to this package's tests, this may
+	// need revisiting to avoid flakes.
+}
+
+// lockedActiveRecordingCount returns the number of recordings that
+// are still actively running. Must be called while p.mu is held.
+// The max concurrency is low (maxConcurrentRecordings = 5), so a
+// full scan is cheap and avoids maintaining a separate counter.
+func (p *portableDesktop) lockedActiveRecordingCount() int {
+	active := 0
+	for _, rec := range p.recordings {
+		if rec.stopped {
+			continue
+		}
+		select {
+		case <-rec.done:
+		default:
+			active++
+		}
+	}
+	return active
+}
+
+// lockedCleanStaleRecordings removes stopped recordings whose temp
+// files are older than one hour. Must be called while p.mu is held.
+func (p *portableDesktop) lockedCleanStaleRecordings(ctx context.Context) {
+	for id, rec := range p.recordings {
+		if !rec.stopped {
+			continue
+		}
+		info, err := os.Stat(rec.filePath)
+		if err != nil {
+			// File already removed or inaccessible; drop entry.
+			delete(p.recordings, id)
+			continue
+		}
+		if p.clock.Since(info.ModTime()) > time.Hour {
+			if err := os.Remove(rec.filePath); err != nil && !os.IsNotExist(err) {
+				p.logger.Warn(ctx, "failed to remove stale recording file",
+					slog.F("recording_id", id),
+					slog.F("file_path", rec.filePath),
+					slog.Error(err),
+				)
+			}
+			delete(p.recordings, id)
+		}
+	}
+}
+
+// Close shuts down the desktop session and cleans up resources.
+func (p *portableDesktop) Close() error {
+	p.mu.Lock()
 	p.closed = true
-	if p.session != nil {
-		p.session.cancel()
-		// Xvnc is a child process — killing it cleans up the X
-		// session.
-		_ = p.session.cmd.Process.Kill()
-		_ = p.session.cmd.Wait()
-		p.session = nil
+
+	// Force-kill all active recordings. The stopOnce inside
+	// lockedStopRecordingProcess makes this safe for
+	// already-stopped recordings.
+	for _, rec := range p.recordings {
+		p.lockedStopRecordingProcess(context.Background(), rec, true)
+	}
+
+	// Snapshot recording file paths and idle goroutine channels
+	// for cleanup, then clear the map.
+	type recEntry struct {
+		id       string
+		filePath string
+		idleDone chan struct{}
+	}
+	var allRecs []recEntry
+	for id, rec := range p.recordings {
+		allRecs = append(allRecs, recEntry{id: id, filePath: rec.filePath, idleDone: rec.idleDone})
+		delete(p.recordings, id)
+	}
+	session := p.session
+	p.session = nil
+	p.mu.Unlock()
+
+	// Wait for all per-recording idle goroutines to exit.
+	for _, entry := range allRecs {
+		if entry.idleDone != nil {
+			<-entry.idleDone
+		}
+	}
+
+	// Remove all recording files and wait for the session to
+	// exit with a timeout so a slow filesystem or hung process
+	// cannot block agent shutdown indefinitely.
+	cleanupDone := make(chan struct{})
+	go func() {
+		defer close(cleanupDone)
+		for _, entry := range allRecs {
+			if err := os.Remove(entry.filePath); err != nil && !os.IsNotExist(err) {
+				p.logger.Warn(context.Background(), "failed to remove recording file on close",
+					slog.F("recording_id", entry.id),
+					slog.F("file_path", entry.filePath),
+					slog.Error(err),
+				)
+			}
+		}
+		if session != nil {
+			session.cancel()
+			if err := session.cmd.Process.Kill(); err != nil {
+				p.logger.Warn(context.Background(), "failed to kill portabledesktop process",
+					slog.Error(err),
+				)
+			}
+			if err := session.cmd.Wait(); err != nil {
+				var exitErr *exec.ExitError
+				if !errors.As(err, &exitErr) {
+					p.logger.Warn(context.Background(), "portabledesktop process exited with error",
+						slog.Error(err),
+					)
+				}
+			}
+		}
+	}()
+	timer := p.clock.NewTimer(15*time.Second, "agentdesktop", "close_cleanup_timeout")
+	defer timer.Stop()
+	select {
+	case <-cleanupDone:
+	case <-timer.C:
+		p.logger.Warn(context.Background(), "timed out waiting for close cleanup")
 	}
 	return nil
 }
 
+// RecordActivity marks the desktop as having received user
+// interaction, resetting the idle-recording timer.
+func (p *portableDesktop) RecordActivity() {
+	p.lastDesktopActionAt.Store(p.clock.Now().UnixNano())
+}
+
 // runCmd executes a portabledesktop subcommand and returns combined
 // output. The caller must have previously called ensureBinary.
 func (p *portableDesktop) runCmd(ctx context.Context, args ...string) (string, error) {
@@ -397,3 +738,31 @@ func (p *portableDesktop) ensureBinary(ctx context.Context) error {
 
 	return xerrors.New("portabledesktop binary not found in PATH or script bin directory")
 }
+
+// monitorRecordingIdle watches for desktop inactivity and stops the
+// given recording when the idle timeout is reached.
+func (p *portableDesktop) monitorRecordingIdle(ctx context.Context, rec *recordingProcess) {
+	timer := p.clock.NewTimer(idleTimeout, "agentdesktop", "recording_idle")
+	defer timer.Stop()
+
+	for {
+		select {
+		case <-timer.C:
+			lastNano := p.lastDesktopActionAt.Load()
+			lastAction := time.Unix(0, lastNano)
+			elapsed := p.clock.Since(lastAction)
+			if elapsed >= idleTimeout {
+				p.mu.Lock()
+				p.lockedStopRecordingProcess(context.Background(), rec, false)
+				p.mu.Unlock()
+				return
+			}
+			// Activity happened; reset with remaining budget.
+			timer.Reset(idleTimeout-elapsed, "agentdesktop", "recording_idle")
+		case <-rec.done:
+			return
+		case <-ctx.Done():
+			return
+		}
+	}
+}

agent/x/agentdesktop/portabledesktop_internal_test.go

@@ -9,13 +9,17 @@ import (
 	"strings"
 	"sync"
 	"testing"
+	"time"
 
+	"github.com/google/uuid"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 
 	"cdr.dev/slog/v3/sloggers/slogtest"
 	"github.com/coder/coder/v2/agent/agentexec"
 	"github.com/coder/coder/v2/pty"
+	"github.com/coder/coder/v2/testutil"
+	"github.com/coder/quartz"
 )
 
 // recordedExecer implements agentexec.Execer by recording every
@@ -86,6 +90,7 @@ func TestPortableDesktop_Start_ParsesOutput(t *testing.T) {
 		execer:       rec,
 		scriptBinDir: t.TempDir(),
 		binPath:      "portabledesktop", // pre-set so ensureBinary is a no-op
+		clock:        quartz.NewReal(),
 	}
 
 	ctx := t.Context()
@@ -117,6 +122,7 @@ func TestPortableDesktop_Start_Idempotent(t *testing.T) {
 		execer:       rec,
 		scriptBinDir: t.TempDir(),
 		binPath:      "portabledesktop",
+		clock:        quartz.NewReal(),
 	}
 
 	ctx := t.Context()
@@ -159,6 +165,7 @@ func TestPortableDesktop_Screenshot(t *testing.T) {
 		execer:       rec,
 		scriptBinDir: t.TempDir(),
 		binPath:      "portabledesktop",
+		clock:        quartz.NewReal(),
 	}
 
 	ctx := t.Context()
@@ -184,6 +191,7 @@ func TestPortableDesktop_Screenshot_WithTargetDimensions(t *testing.T) {
 		execer:       rec,
 		scriptBinDir: t.TempDir(),
 		binPath:      "portabledesktop",
+		clock:        quartz.NewReal(),
 	}
 
 	ctx := t.Context()
@@ -282,6 +290,7 @@ func TestPortableDesktop_MouseMethods(t *testing.T) {
 				execer:       rec,
 				scriptBinDir: t.TempDir(),
 				binPath:      "portabledesktop",
+				clock:        quartz.NewReal(),
 			}
 
 			err := tt.invoke(t.Context(), pd)
@@ -289,7 +298,6 @@ func TestPortableDesktop_MouseMethods(t *testing.T) {
 
 			cmds := rec.allCommands()
 			require.NotEmpty(t, cmds, "expected at least one command")
-
 			// Find at least one recorded command that contains
 			// all expected argument substrings.
 			found := false
@@ -367,6 +375,7 @@ func TestPortableDesktop_KeyboardMethods(t *testing.T) {
 				execer:       rec,
 				scriptBinDir: t.TempDir(),
 				binPath:      "portabledesktop",
+				clock:        quartz.NewReal(),
 			}
 
 			err := tt.invoke(t.Context(), pd)
@@ -423,6 +432,7 @@ func TestPortableDesktop_Close(t *testing.T) {
 		execer:       rec,
 		scriptBinDir: t.TempDir(),
 		binPath:      "portabledesktop",
+		clock:        quartz.NewReal(),
 	}
 
 	ctx := t.Context()
@@ -445,7 +455,7 @@ func TestPortableDesktop_Close(t *testing.T) {
 	// Subsequent Start must fail.
 	_, err = pd.Start(ctx)
 	require.Error(t, err)
-	assert.Contains(t, err.Error(), "desktop is closed")
+	assert.Contains(t, err.Error(), "desktop closed")
 }
 
 // --- ensureBinary tests ---
@@ -539,7 +549,410 @@ func TestEnsureBinary_NotFound(t *testing.T) {
 	assert.Contains(t, err.Error(), "not found")
 }
 
+func TestPortableDesktop_StartRecording(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"record": `trap 'exit 0' INT; sleep 120 & wait`,
+			"up":     `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	clk := quartz.NewReal()
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	ctx := t.Context()
+	recID := uuid.New().String()
+	err := pd.StartRecording(ctx, recID)
+	require.NoError(t, err)
+
+	cmds := rec.allCommands()
+	require.NotEmpty(t, cmds)
+	// Find the record command (not the up command).
+	found := false
+	for _, cmd := range cmds {
+		joined := strings.Join(cmd, " ")
+		if strings.Contains(joined, "record") && strings.Contains(joined, "coder-recording-"+recID) {
+			found = true
+			break
+		}
+	}
+	assert.True(t, found, "expected a record command with the recording ID")
+
+	require.NoError(t, pd.Close())
+}
+
+func TestPortableDesktop_StartRecording_ConcurrentLimit(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"record": `trap 'exit 0' INT; sleep 120 & wait`,
+			"up":     `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	clk := quartz.NewReal()
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	ctx := t.Context()
+
+	for i := range maxConcurrentRecordings {
+		err := pd.StartRecording(ctx, uuid.New().String())
+		require.NoError(t, err, "recording %d should succeed", i)
+	}
+
+	err := pd.StartRecording(ctx, uuid.New().String())
+	require.Error(t, err)
+	assert.Contains(t, err.Error(), "too many concurrent recordings")
+
+	require.NoError(t, pd.Close())
+}
+
+func TestPortableDesktop_StopRecording_ReturnsArtifact(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"record": `trap 'exit 0' INT; sleep 120 & wait`,
+			"up":     `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	clk := quartz.NewReal()
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	ctx := t.Context()
+	recID := uuid.New().String()
+	err := pd.StartRecording(ctx, recID)
+	require.NoError(t, err)
+
+	// Write a dummy MP4 file at the expected path so StopRecording
+	// can open it as an artifact.
+	filePath := filepath.Join(os.TempDir(), "coder-recording-"+recID+".mp4")
+	require.NoError(t, os.WriteFile(filePath, []byte("fake-mp4-data"), 0o600))
+	t.Cleanup(func() { _ = os.Remove(filePath) })
+
+	artifact, err := pd.StopRecording(ctx, recID)
+	require.NoError(t, err)
+	defer artifact.Reader.Close()
+	assert.Equal(t, int64(len("fake-mp4-data")), artifact.Size)
+
+	require.NoError(t, pd.Close())
+}
+
+func TestPortableDesktop_StopRecording_UnknownID(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"record": `trap 'exit 0' INT; sleep 120 & wait`,
+		},
+	}
+
+	clk := quartz.NewReal()
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	ctx := t.Context()
+	_, err := pd.StopRecording(ctx, uuid.New().String())
+	require.ErrorIs(t, err, ErrUnknownRecording)
+
+	require.NoError(t, pd.Close())
+}
+
 // Ensure that portableDesktop satisfies the Desktop interface at
 // compile time. This uses the unexported type so it lives in the
 // internal test package.
 var _ Desktop = (*portableDesktop)(nil)
+
+func TestPortableDesktop_IdleTimeout_StopsRecordings(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"record": `trap 'exit 0' INT; sleep 120 & wait`,
+			"up":     `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	clk := quartz.NewMock(t)
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	ctx := t.Context()
+	recID := uuid.New().String()
+
+	// Install the trap before StartRecording so it is guaranteed
+	// to catch the idle monitor's NewTimer call regardless of
+	// goroutine scheduling.
+	trap := clk.Trap().NewTimer("agentdesktop", "recording_idle")
+
+	err := pd.StartRecording(ctx, recID)
+	require.NoError(t, err)
+
+	// Verify recording is active.
+	pd.mu.Lock()
+	require.False(t, pd.recordings[recID].stopped)
+	pd.mu.Unlock()
+
+	// Wait for the idle monitor timer to be created and release
+	// it so the monitor enters its select loop.
+	trap.MustWait(ctx).MustRelease(ctx)
+	trap.Close()
+
+	// The stop-all path calls lockedStopRecordingProcess which
+	// creates a per-recording 15s stop_timeout timer.
+	stopTrap := clk.Trap().NewTimer("agentdesktop", "stop_timeout")
+
+	// Advance past idle timeout to trigger the stop-all.
+	clk.Advance(idleTimeout)
+
+	// Wait for the stop timer to be created, then release it.
+	stopTrap.MustWait(ctx).MustRelease(ctx)
+	stopTrap.Close()
+
+	// The recording process should now be stopped.
+	require.Eventually(t, func() bool {
+		pd.mu.Lock()
+		defer pd.mu.Unlock()
+		rec, ok := pd.recordings[recID]
+		return ok && rec.stopped
+	}, testutil.WaitShort, testutil.IntervalFast)
+
+	require.NoError(t, pd.Close())
+}
+
+func TestPortableDesktop_IdleTimeout_ActivityResetsTimer(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"record": `trap 'exit 0' INT; sleep 120 & wait`,
+			"up":     `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	clk := quartz.NewMock(t)
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	ctx := t.Context()
+	recID := uuid.New().String()
+
+	// Install the trap before StartRecording so it is guaranteed
+	// to catch the idle monitor's NewTimer call regardless of
+	// goroutine scheduling.
+	trap := clk.Trap().NewTimer("agentdesktop", "recording_idle")
+
+	err := pd.StartRecording(ctx, recID)
+	require.NoError(t, err)
+
+	// Wait for the idle monitor timer to be created.
+	trap.MustWait(ctx).MustRelease(ctx)
+	trap.Close()
+
+	// Advance most of the way but not past the timeout.
+	clk.Advance(idleTimeout - time.Minute)
+
+	// Record activity to reset the timer.
+	pd.RecordActivity()
+
+	// Trap the Reset call that the idle monitor makes when it
+	// sees recent activity.
+	resetTrap := clk.Trap().TimerReset("agentdesktop", "recording_idle")
+
+	// Advance past the original idle timeout deadline. The
+	// monitor should see the recent activity and reset instead
+	// of stopping.
+	clk.Advance(time.Minute)
+
+	resetTrap.MustWait(ctx).MustRelease(ctx)
+	resetTrap.Close()
+
+	// Recording should still be active because activity was
+	// recorded.
+	pd.mu.Lock()
+	require.False(t, pd.recordings[recID].stopped)
+	pd.mu.Unlock()
+
+	require.NoError(t, pd.Close())
+}
+
+func TestPortableDesktop_IdleTimeout_MultipleRecordings(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"record": `trap 'exit 0' INT; sleep 120 & wait`,
+			"up":     `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	clk := quartz.NewMock(t)
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	ctx := t.Context()
+	recID1 := uuid.New().String()
+	recID2 := uuid.New().String()
+
+	// Trap idle timer creation for both recordings.
+	trap := clk.Trap().NewTimer("agentdesktop", "recording_idle")
+
+	err := pd.StartRecording(ctx, recID1)
+	require.NoError(t, err)
+
+	// Wait for first recording's idle timer.
+	trap.MustWait(ctx).MustRelease(ctx)
+
+	err = pd.StartRecording(ctx, recID2)
+	require.NoError(t, err)
+
+	// Wait for second recording's idle timer.
+	trap.MustWait(ctx).MustRelease(ctx)
+	trap.Close()
+
+	// Trap the stop timers that will be created when idle fires.
+	stopTrap := clk.Trap().NewTimer("agentdesktop", "stop_timeout")
+
+	// Advance past idle timeout.
+	clk.Advance(idleTimeout)
+
+	// Wait for both stop timers.
+	stopTrap.MustWait(ctx).MustRelease(ctx)
+	stopTrap.MustWait(ctx).MustRelease(ctx)
+	stopTrap.Close()
+
+	// Both recordings should be stopped.
+	require.Eventually(t, func() bool {
+		pd.mu.Lock()
+		defer pd.mu.Unlock()
+		r1, ok1 := pd.recordings[recID1]
+		r2, ok2 := pd.recordings[recID2]
+		return ok1 && r1.stopped && ok2 && r2.stopped
+	}, testutil.WaitShort, testutil.IntervalFast)
+
+	require.NoError(t, pd.Close())
+}
+
+func TestPortableDesktop_StartRecording_ReturnsErrDesktopClosed(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"up": `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	clk := quartz.NewReal()
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        clk,
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(clk.Now().UnixNano())
+
+	// Start and close the desktop so it's in the closed state.
+	ctx := t.Context()
+	_, err := pd.Start(ctx)
+	require.NoError(t, err)
+	require.NoError(t, pd.Close())
+
+	// StartRecording should now return ErrDesktopClosed.
+	err = pd.StartRecording(ctx, uuid.New().String())
+	require.ErrorIs(t, err, ErrDesktopClosed)
+}
+
+func TestPortableDesktop_Start_ReturnsErrDesktopClosed(t *testing.T) {
+	t.Parallel()
+
+	logger := slogtest.Make(t, nil)
+	rec := &recordedExecer{
+		scripts: map[string]string{
+			"up": `printf '{"vncPort":5901,"geometry":"1920x1080"}\n' && sleep 120`,
+		},
+	}
+
+	pd := &portableDesktop{
+		logger:       logger,
+		execer:       rec,
+		scriptBinDir: t.TempDir(),
+		clock:        quartz.NewReal(),
+		binPath:      "portabledesktop",
+		recordings:   make(map[string]*recordingProcess),
+	}
+	pd.lastDesktopActionAt.Store(pd.clock.Now().UnixNano())
+
+	ctx := t.Context()
+	_, err := pd.Start(ctx)
+	require.NoError(t, err)
+	require.NoError(t, pd.Close())
+
+	_, err = pd.Start(ctx)
+	require.ErrorIs(t, err, ErrDesktopClosed)
+}

agent/x/agentdesktop/portabledesktop_stop_other.go

@@ -0,0 +1,12 @@
+//go:build !windows
+
+package agentdesktop
+
+import "os"
+
+// interruptRecordingProcess sends a SIGINT to the recording process
+// for graceful shutdown. On Unix, os.Interrupt is delivered as
+// SIGINT which lets the recorder finalize the MP4 container.
+func interruptRecordingProcess(p *os.Process) error {
+	return p.Signal(os.Interrupt)
+}

agent/x/agentdesktop/portabledesktop_stop_windows.go

@@ -0,0 +1,10 @@
+package agentdesktop
+
+import "os"
+
+// interruptRecordingProcess kills the recording process directly
+// because os.Process.Signal(os.Interrupt) is not supported on
+// Windows and returns an error without delivering a signal.
+func interruptRecordingProcess(p *os.Process) error {
+	return p.Kill()
+}

agent/x/agentmcp/manager.go

@@ -6,7 +6,6 @@ import (
 	"fmt"
 	"io/fs"
 	"os"
-	"path/filepath"
 	"slices"
 	"strings"
 	"sync"
@@ -70,16 +69,40 @@ func NewManager(logger slog.Logger) *Manager {
 	}
 }
 
-// Connect discovers .mcp.json in dir and connects to all
-// configured servers. Failed servers are logged and skipped.
-func (m *Manager) Connect(ctx context.Context, dir string) error {
-	path := filepath.Join(dir, ".mcp.json")
-	configs, err := ParseConfig(path)
-	if err != nil {
-		if errors.Is(err, fs.ErrNotExist) {
-			return nil
+// Connect reads MCP config files at the given absolute paths and
+// connects to all configured servers. Failed servers are logged
+// and skipped. Missing config files are silently skipped.
+func (m *Manager) Connect(ctx context.Context, mcpConfigFiles []string) error {
+	var allConfigs []ServerConfig
+	for _, configPath := range mcpConfigFiles {
+		configs, err := ParseConfig(configPath)
+		if err != nil {
+			if errors.Is(err, fs.ErrNotExist) {
+				continue
+			}
+			m.logger.Warn(ctx, "failed to parse MCP config",
+				slog.F("path", configPath),
+				slog.Error(err),
+			)
+			continue
 		}
-		return xerrors.Errorf("parse mcp config: %w", err)
+		allConfigs = append(allConfigs, configs...)
+	}
+
+	// Deduplicate by server name; first occurrence wins.
+	seen := make(map[string]struct{})
+	deduped := make([]ServerConfig, 0, len(allConfigs))
+	for _, cfg := range allConfigs {
+		if _, ok := seen[cfg.Name]; ok {
+			continue
+		}
+		seen[cfg.Name] = struct{}{}
+		deduped = append(deduped, cfg)
+	}
+	allConfigs = deduped
+
+	if len(allConfigs) == 0 {
+		return nil
 	}
 
 	// Connect to servers in parallel without holding the
@@ -95,7 +118,7 @@ func (m *Manager) Connect(ctx context.Context, dir string) error {
 		connected []connectedServer
 	)
 	var eg errgroup.Group
-	for _, cfg := range configs {
+	for _, cfg := range allConfigs {
 		eg.Go(func() error {
 			c, err := m.connectServer(ctx, cfg)
 			if err != nil {
@@ -164,7 +187,11 @@ func (*Manager) connectServer(ctx context.Context, cfg ServerConfig) (*client.Cl
 	connectCtx, cancel := context.WithTimeout(ctx, connectTimeout)
 	defer cancel()
 
-	if err := c.Start(connectCtx); err != nil {
+	// Use the parent ctx (not connectCtx) so the subprocess outlives
+	// the connect/initialize handshake. connectCtx bounds only the
+	// Initialize call below. The subprocess is cleaned up when the
+	// Manager is closed or ctx is canceled.
+	if err := c.Start(ctx); err != nil {
 		_ = c.Close()
 		return nil, xerrors.Errorf("start %q: %w", cfg.Name, err)
 	}

agent/x/agentmcp/manager_internal_test.go

@@ -1,6 +1,11 @@
 package agentmcp
 
 import (
+	"bufio"
+	"context"
+	"encoding/json"
+	"fmt"
+	"os"
 	"testing"
 
 	"github.com/mark3labs/mcp-go/mcp"
@@ -8,6 +13,7 @@ import (
 	"github.com/stretchr/testify/require"
 
 	"github.com/coder/coder/v2/codersdk/workspacesdk"
+	"github.com/coder/coder/v2/testutil"
 )
 
 func TestSplitToolName(t *testing.T) {
@@ -193,3 +199,118 @@ func TestConvertResult(t *testing.T) {
 		})
 	}
 }
+
+// TestConnectServer_StdioProcessSurvivesConnect verifies that a stdio MCP
+// server subprocess remains alive after connectServer returns. This is a
+// regression test for a bug where the subprocess was tied to a short-lived
+// connectCtx and killed as soon as the context was canceled.
+func TestConnectServer_StdioProcessSurvivesConnect(t *testing.T) {
+	t.Parallel()
+
+	if os.Getenv("TEST_MCP_FAKE_SERVER") == "1" {
+		// Child process: act as a minimal MCP server over stdio.
+		runFakeMCPServer()
+		return
+	}
+
+	// Get the path to the test binary so we can re-exec ourselves
+	// as a fake MCP server subprocess.
+	testBin, err := os.Executable()
+	require.NoError(t, err)
+
+	cfg := ServerConfig{
+		Name:      "fake",
+		Transport: "stdio",
+		Command:   testBin,
+		Args:      []string{"-test.run=^TestConnectServer_StdioProcessSurvivesConnect$"},
+		Env:       map[string]string{"TEST_MCP_FAKE_SERVER": "1"},
+	}
+
+	ctx := testutil.Context(t, testutil.WaitLong)
+	m := &Manager{}
+	client, err := m.connectServer(ctx, cfg)
+	require.NoError(t, err, "connectServer should succeed")
+	t.Cleanup(func() { _ = client.Close() })
+
+	// At this point connectServer has returned and its internal
+	// connectCtx has been canceled. The subprocess must still be
+	// alive. Verify by listing tools (requires a live server).
+	listCtx, listCancel := context.WithTimeout(ctx, testutil.WaitShort)
+	defer listCancel()
+	result, err := client.ListTools(listCtx, mcp.ListToolsRequest{})
+	require.NoError(t, err, "ListTools should succeed — server must be alive after connect")
+	require.Len(t, result.Tools, 1)
+	assert.Equal(t, "echo", result.Tools[0].Name)
+}
+
+// runFakeMCPServer implements a minimal JSON-RPC / MCP server over
+// stdin/stdout, just enough for initialize + tools/list.
+func runFakeMCPServer() {
+	scanner := bufio.NewScanner(os.Stdin)
+	for scanner.Scan() {
+		line := scanner.Bytes()
+
+		var req struct {
+			JSONRPC string          `json:"jsonrpc"`
+			ID      json.RawMessage `json:"id"`
+			Method  string          `json:"method"`
+		}
+		if err := json.Unmarshal(line, &req); err != nil {
+			continue
+		}
+
+		var resp any
+		switch req.Method {
+		case "initialize":
+			resp = map[string]any{
+				"jsonrpc": "2.0",
+				"id":      req.ID,
+				"result": map[string]any{
+					"protocolVersion": "2025-03-26",
+					"capabilities": map[string]any{
+						"tools": map[string]any{},
+					},
+					"serverInfo": map[string]any{
+						"name":    "fake-server",
+						"version": "0.0.1",
+					},
+				},
+			}
+		case "notifications/initialized":
+			// No response needed for notifications.
+			continue
+		case "tools/list":
+			resp = map[string]any{
+				"jsonrpc": "2.0",
+				"id":      req.ID,
+				"result": map[string]any{
+					"tools": []map[string]any{
+						{
+							"name":        "echo",
+							"description": "echoes input",
+							"inputSchema": map[string]any{
+								"type":       "object",
+								"properties": map[string]any{},
+							},
+						},
+					},
+				},
+			}
+		default:
+			resp = map[string]any{
+				"jsonrpc": "2.0",
+				"id":      req.ID,
+				"error": map[string]any{
+					"code":    -32601,
+					"message": "method not found",
+				},
+			}
+		}
+
+		out, err := json.Marshal(resp)
+		if err != nil {
+			continue
+		}
+		_, _ = fmt.Fprintf(os.Stdout, "%s\n", out)
+	}
+}

biome.jsonc

@@ -3,11 +3,13 @@
 		"enabled": true,
 		"clientKind": "git",
 		"useIgnoreFile": true,
-		"defaultBranch": "main",
+		"defaultBranch": "main"
 	},
 	"files": {
-		"includes": ["**", "!**/pnpm-lock.yaml"],
-		"ignoreUnknown": true,
+		// static/*.html are Go templates with {{ }} directives that
+		// Biome's HTML parser does not support.
+		"includes": ["**", "!**/pnpm-lock.yaml", "!**/static/*.html"],
+		"ignoreUnknown": true
 	},
 	"linter": {
 		"rules": {
@@ -15,7 +17,7 @@
 				"noSvgWithoutTitle": "off",
 				"useButtonType": "off",
 				"useSemanticElements": "off",
-				"noStaticElementInteractions": "off",
+				"noStaticElementInteractions": "off"
 			},
 			"correctness": {
 				"noUnusedImports": "warn",
@@ -24,9 +26,9 @@
 				"noUnusedVariables": {
 					"level": "warn",
 					"options": {
-						"ignoreRestSiblings": true,
-					},
-				},
+						"ignoreRestSiblings": true
+					}
+				}
 			},
 			"style": {
 				"noNonNullAssertion": "off",
@@ -47,7 +49,7 @@
 						"paths": {
 							"react": {
 								"message": "React 19 no longer requires forwardRef. Use ref as a prop instead.",
-								"importNames": ["forwardRef"],
+								"importNames": ["forwardRef"]
 							},
 							// "@mui/material/Alert": "Use components/Alert/Alert instead.",
 							// "@mui/material/AlertTitle": "Use components/Alert/Alert instead.",
@@ -115,10 +117,10 @@
 							"@emotion/styled": "Use Tailwind CSS instead.",
 							// "@emotion/cache": "Use Tailwind CSS instead.",
 							// "components/Stack/Stack": "Use Tailwind flex utilities instead (e.g., <div className='flex flex-col gap-4'>).",
-							"lodash": "Use lodash/<name> instead.",
-						},
-					},
-				},
+							"lodash": "Use lodash/<name> instead."
+						}
+					}
+				}
 			},
 			"suspicious": {
 				"noArrayIndexKey": "off",
@@ -129,14 +131,21 @@
 				"noConsole": {
 					"level": "error",
 					"options": {
-						"allow": ["error", "info", "warn"],
-					},
-				},
+						"allow": ["error", "info", "warn"]
+					}
+				}
 			},
 			"complexity": {
-				"noImportantStyles": "off", // TODO: check and fix !important styles
-			},
-		},
+				"noImportantStyles": "off" // TODO: check and fix !important styles
+			}
+		}
 	},
-	"$schema": "./node_modules/@biomejs/biome/configuration_schema.json",
+	"css": {
+		"parser": {
+			// Biome 2.3+ requires opt-in for @apply and other
+			// Tailwind directives.
+			"tailwindDirectives": true
+		}
+	},
+	"$schema": "./node_modules/@biomejs/biome/configuration_schema.json"
 }

cli/agent.go

@@ -17,6 +17,7 @@ import (
 	"strings"
 	"time"
 
+	"github.com/google/uuid"
 	"github.com/prometheus/client_golang/prometheus"
 	"golang.org/x/xerrors"
 	"gopkg.in/natefinch/lumberjack.v2"
@@ -272,11 +273,14 @@ func workspaceAgent() *serpent.Command {
 				logger.Info(ctx, "agent devcontainer detection not enabled")
 			}
 
-			reinitEvents := agentsdk.WaitForReinitLoop(ctx, logger, client)
+			reinitCtx, reinitCancel := context.WithCancel(ctx)
+			defer reinitCancel()
+			reinitEvents := agentsdk.WaitForReinitLoop(reinitCtx, logger, client)
 
 			var (
-				lastErr  error
-				mustExit bool
+				lastOwnerID uuid.UUID
+				lastErr     error
+				mustExit    bool
 			)
 			for {
 				prometheusRegistry := prometheus.NewRegistry()
@@ -343,9 +347,32 @@ func workspaceAgent() *serpent.Command {
 				case <-ctx.Done():
 					logger.Info(ctx, "agent shutting down", slog.Error(context.Cause(ctx)))
 					mustExit = true
-				case event := <-reinitEvents:
-					logger.Info(ctx, "agent received instruction to reinitialize",
-						slog.F("workspace_id", event.WorkspaceID), slog.F("reason", event.Reason))
+				case event, ok := <-reinitEvents:
+					switch {
+					case !ok:
+						// Channel closed — the reinit loop exited
+						// (terminal 409 or context expired). Keep
+						// running the current agent until the parent
+						// context is canceled.
+						logger.Info(ctx, "reinit channel closed, running without reinit capability")
+						reinitEvents = nil
+						<-ctx.Done()
+						mustExit = true
+					case event.OwnerID != uuid.Nil && event.OwnerID == lastOwnerID:
+						// Duplicate reinit for same owner — already
+						// reinitialized. Cancel the reinit loop
+						// goroutine and keep the current agent.
+						logger.Info(ctx, "skipping redundant reinit, owner unchanged",
+							slog.F("owner_id", event.OwnerID))
+						reinitCancel()
+						reinitEvents = nil
+						<-ctx.Done()
+						mustExit = true
+					default:
+						lastOwnerID = event.OwnerID
+						logger.Info(ctx, "agent received instruction to reinitialize",
+							slog.F("workspace_id", event.WorkspaceID), slog.F("reason", event.Reason))
+					}
 				}
 
 				lastErr = agnt.Close()

cli/clilog/clilog.go

@@ -104,7 +104,7 @@ func (b *Builder) Build(inv *serpent.Invocation) (log slog.Logger, closeLog func
 
 	addSinkIfProvided := func(sinkFn func(io.Writer) slog.Sink, loc string) error {
 		switch loc {
-		case "":
+		case "", "/dev/null":
 		case "/dev/stdout":
 			sinks = append(sinks, sinkFn(inv.Stdout))
 

cli/exp_scaletest.go

@@ -1401,6 +1401,9 @@ func (r *RootCmd) scaletestWorkspaceTraffic() *serpent.Command {
 				// Setup our workspace agent connection.
 				config := workspacetraffic.Config{
 					AgentID:       agent.ID,
+					WorkspaceID:   ws.ID,
+					WorkspaceName: ws.Name,
+					AgentName:     agent.Name,
 					BytesPerTick:  bytesPerTick,
 					Duration:      strategy.timeout,
 					TickInterval:  tickInterval,

cli/schedule_test.go

@@ -352,8 +352,6 @@ func TestScheduleOverride(t *testing.T) {
 			require.NoError(t, err, "invalid schedule")
 			ownerClient, _, _, ws := setupTestSchedule(t, sched)
 			now := time.Now()
-			// To avoid the likelihood of time-related flakes, only matching up to the hour.
-			expectedDeadline := now.In(loc).Add(10 * time.Hour).Format("2006-01-02T15:")
 
 			// When: we override the stop schedule
 			inv, root := clitest.New(t,
@@ -364,6 +362,19 @@ func TestScheduleOverride(t *testing.T) {
 			pty := ptytest.New(t).Attach(inv)
 			require.NoError(t, inv.Run())
 
+			// Fetch the workspace to get the actual deadline set by the
+			// server. Computing our own expected deadline from a separately
+			// captured time.Now() is racy: the CLI command calls time.Now()
+			// internally, and with the Asia/Kolkata +05:30 offset the hour
+			// boundary falls at :30 UTC minutes. A small delay between our
+			// time.Now() and the command's is enough to land in different
+			// hours.
+			updated, err := ownerClient.Workspace(context.Background(), ws[0].ID)
+			require.NoError(t, err)
+			require.False(t, updated.LatestBuild.Deadline.IsZero(), "deadline should be set after extend")
+			require.WithinDuration(t, now.Add(10*time.Hour), updated.LatestBuild.Deadline.Time, 5*time.Minute)
+			expectedDeadline := updated.LatestBuild.Deadline.Time.In(loc).Format(time.RFC3339)
+
 			// Then: the updated schedule should be shown
 			pty.ExpectMatch(ws[0].OwnerName + "/" + ws[0].Name)
 			pty.ExpectMatch(sched.Humanize())

cli/sync_test.go

@@ -165,6 +165,37 @@ func TestSyncCommands_Golden(t *testing.T) {
 		clitest.TestGoldenFile(t, "TestSyncCommands_Golden/want_success", outBuf.Bytes(), nil)
 	})
 
+	t.Run("want_multiple_deps", func(t *testing.T) {
+		t.Parallel()
+		path, cleanup := setupSocketServer(t)
+		defer cleanup()
+
+		ctx := testutil.Context(t, testutil.WaitShort)
+
+		var outBuf bytes.Buffer
+		inv, _ := clitest.New(t, "exp", "sync", "want", "test-unit", "dep-1", "dep-2", "dep-3", "--socket-path", path)
+		inv.Stdout = &outBuf
+		inv.Stderr = &outBuf
+
+		err := inv.WithContext(ctx).Run()
+		require.NoError(t, err)
+
+		// Verify all dependencies were registered by checking status.
+		outBuf.Reset()
+		inv, _ = clitest.New(t, "exp", "sync", "status", "test-unit", "--socket-path", path, "--output", "json")
+		inv.Stdout = &outBuf
+		inv.Stderr = &outBuf
+
+		err = inv.WithContext(ctx).Run()
+		require.NoError(t, err)
+
+		// The output should mention all three dependencies.
+		output := outBuf.String()
+		require.Contains(t, output, "dep-1")
+		require.Contains(t, output, "dep-2")
+		require.Contains(t, output, "dep-3")
+	})
+
 	t.Run("complete", func(t *testing.T) {
 		t.Parallel()
 		path, cleanup := setupSocketServer(t)

cli/sync_want.go

@@ -11,17 +11,16 @@ import (
 
 func (*RootCmd) syncWant(socketPath *string) *serpent.Command {
 	cmd := &serpent.Command{
-		Use:   "want <unit> <depends-on>",
-		Short: "Declare that a unit depends on another unit completing before it can start",
-		Long:  "Declare that a unit depends on another unit completing before it can start. The unit specified first will not start until the second has signaled that it has completed.",
+		Use:   "want <unit> <depends-on> [depends-on...]",
+		Short: "Declare that a unit depends on other units completing before it can start",
+		Long:  "Declare that a unit depends on one or more other units completing before it can start. The unit specified first will not start until all subsequent units have signaled that they have completed.",
 		Handler: func(i *serpent.Invocation) error {
 			ctx := i.Context()
 
-			if len(i.Args) != 2 {
-				return xerrors.New("exactly two arguments are required: unit and depends-on")
+			if len(i.Args) < 2 {
+				return xerrors.New("at least two arguments are required: unit and one or more depends-on")
 			}
 			dependentUnit := unit.ID(i.Args[0])
-			dependsOn := unit.ID(i.Args[1])
 
 			opts := []agentsocket.Option{}
 			if *socketPath != "" {
@@ -34,8 +33,10 @@ func (*RootCmd) syncWant(socketPath *string) *serpent.Command {
 			}
 			defer client.Close()
 
-			if err := client.SyncWant(ctx, dependentUnit, dependsOn); err != nil {
-				return xerrors.Errorf("declare dependency failed: %w", err)
+			for _, dep := range i.Args[1:] {
+				if err := client.SyncWant(ctx, dependentUnit, unit.ID(dep)); err != nil {
+					return xerrors.Errorf("declare dependency failed: %w", err)
+				}
 			}
 
 			cliui.Info(i.Stdout, "Success")

cli/testdata/coder_exp_sync_--help.golden

@@ -16,7 +16,7 @@ SUBCOMMANDS:
     ping        Test agent socket connectivity and health
     start       Wait until all unit dependencies are satisfied
     status      Show unit status and dependency state
-    want        Declare that a unit depends on another unit completing before it
+    want        Declare that a unit depends on other units completing before it
                 can start
 
 OPTIONS:

cli/testdata/coder_exp_sync_want_--help.golden

@@ -1,13 +1,13 @@
 coder v0.0.0-devel
 
 USAGE:
-  coder exp sync want <unit> <depends-on>
+  coder exp sync want <unit> <depends-on> [depends-on...]
 
-  Declare that a unit depends on another unit completing before it can start
+  Declare that a unit depends on other units completing before it can start
 
-  Declare that a unit depends on another unit completing before it can start.
-  The unit specified first will not start until the second has signaled that it
-  has completed.
+  Declare that a unit depends on one or more other units completing before it
+  can start. The unit specified first will not start until all subsequent units
+  have signaled that they have completed.
 
 ———
 Run `coder --help` for a list of global options.

coderd/agentapi/connectionlog.go

@@ -85,7 +85,7 @@ func (a *ConnLogAPI) ReportConnection(ctx context.Context, req *agentproto.Repor
 		AgentName:        a.AgentName,
 		Type:             connectionType,
 		Code:             code,
-		Ip:               logIP,
+		IP:               logIP,
 		ConnectionID: uuid.NullUUID{
 			UUID:  connectionID,
 			Valid: true,

coderd/agentapi/connectionlog_test.go

@@ -152,7 +152,7 @@ func TestConnectionLog(t *testing.T) {
 					Int32: tt.status,
 					Valid: *tt.action == agentproto.Connection_DISCONNECT,
 				},
-				Ip:   expectedIP,
+				IP:   expectedIP,
 				Type: agentProtoConnectionTypeToConnectionLog(t, *tt.typ),
 				DisconnectReason: sql.NullString{
 					String: tt.reason,

coderd/apidoc/docs.go

@@ -10205,12 +10205,26 @@ const docTemplate = `{
                 ],
                 "summary": "Get workspace agent reinitialization",
                 "operationId": "get-workspace-agent-reinitialization",
+                "parameters": [
+                    {
+                        "type": "boolean",
+                        "description": "Opt in to durable reinit checks",
+                        "name": "wait",
+                        "in": "query"
+                    }
+                ],
                 "responses": {
                     "200": {
                         "description": "OK",
                         "schema": {
                             "$ref": "#/definitions/agentsdk.ReinitializationEvent"
                         }
+                    },
+                    "409": {
+                        "description": "Conflict",
+                        "schema": {
+                            "$ref": "#/definitions/codersdk.Response"
+                        }
                     }
                 },
                 "security": [
@@ -12647,11 +12661,16 @@ const docTemplate = `{
         "agentsdk.ReinitializationEvent": {
             "type": "object",
             "properties": {
+                "owner_id": {
+                    "type": "string",
+                    "format": "uuid"
+                },
                 "reason": {
                     "$ref": "#/definitions/agentsdk.ReinitializationReason"
                 },
-                "workspaceID": {
-                    "type": "string"
+                "workspace_id": {
+                    "type": "string",
+                    "format": "uuid"
                 }
             }
         },
@@ -12894,6 +12913,9 @@ const docTemplate = `{
                 "provider": {
                     "type": "string"
                 },
+                "provider_name": {
+                    "type": "string"
+                },
                 "started_at": {
                     "type": "string",
                     "format": "date-time"
@@ -13111,6 +13133,12 @@ const docTemplate = `{
         "codersdk.AIBridgeSessionThreadsTokenUsage": {
             "type": "object",
             "properties": {
+                "cache_read_input_tokens": {
+                    "type": "integer"
+                },
+                "cache_write_input_tokens": {
+                    "type": "integer"
+                },
                 "input_tokens": {
                     "type": "integer"
                 },
@@ -13126,6 +13154,12 @@ const docTemplate = `{
         "codersdk.AIBridgeSessionTokenUsageSummary": {
             "type": "object",
             "properties": {
+                "cache_read_input_tokens": {
+                    "type": "integer"
+                },
+                "cache_write_input_tokens": {
+                    "type": "integer"
+                },
                 "input_tokens": {
                     "type": "integer"
                 },
@@ -13172,6 +13206,12 @@ const docTemplate = `{
         "codersdk.AIBridgeTokenUsage": {
             "type": "object",
             "properties": {
+                "cache_read_input_tokens": {
+                    "type": "integer"
+                },
+                "cache_write_input_tokens": {
+                    "type": "integer"
+                },
                 "created_at": {
                     "type": "string",
                     "format": "date-time"
@@ -14135,6 +14175,9 @@ const docTemplate = `{
                 },
                 "count": {
                     "type": "integer"
+                },
+                "count_cap": {
+                    "type": "integer"
                 }
             }
         },
@@ -14456,6 +14499,9 @@ const docTemplate = `{
                 },
                 "count": {
                     "type": "integer"
+                },
+                "count_cap": {
+                    "type": "integer"
                 }
             }
         },
@@ -14543,6 +14589,17 @@ const docTemplate = `{
                 }
             }
         },
+        "codersdk.CreateFirstUserOnboardingInfo": {
+            "type": "object",
+            "properties": {
+                "newsletter_marketing": {
+                    "type": "boolean"
+                },
+                "newsletter_releases": {
+                    "type": "boolean"
+                }
+            }
+        },
         "codersdk.CreateFirstUserRequest": {
             "type": "object",
             "required": [
@@ -14557,6 +14614,9 @@ const docTemplate = `{
                 "name": {
                     "type": "string"
                 },
+                "onboarding_info": {
+                    "$ref": "#/definitions/codersdk.CreateFirstUserOnboardingInfo"
+                },
                 "password": {
                     "type": "string"
                 },

coderd/apidoc/swagger.json

@@ -9038,12 +9038,26 @@
 				"tags": ["Agents"],
 				"summary": "Get workspace agent reinitialization",
 				"operationId": "get-workspace-agent-reinitialization",
+				"parameters": [
+					{
+						"type": "boolean",
+						"description": "Opt in to durable reinit checks",
+						"name": "wait",
+						"in": "query"
+					}
+				],
 				"responses": {
 					"200": {
 						"description": "OK",
 						"schema": {
 							"$ref": "#/definitions/agentsdk.ReinitializationEvent"
 						}
+					},
+					"409": {
+						"description": "Conflict",
+						"schema": {
+							"$ref": "#/definitions/codersdk.Response"
+						}
 					}
 				},
 				"security": [
@@ -11229,11 +11243,16 @@
 		"agentsdk.ReinitializationEvent": {
 			"type": "object",
 			"properties": {
+				"owner_id": {
+					"type": "string",
+					"format": "uuid"
+				},
 				"reason": {
 					"$ref": "#/definitions/agentsdk.ReinitializationReason"
 				},
-				"workspaceID": {
-					"type": "string"
+				"workspace_id": {
+					"type": "string",
+					"format": "uuid"
 				}
 			}
 		},
@@ -11472,6 +11491,9 @@
 				"provider": {
 					"type": "string"
 				},
+				"provider_name": {
+					"type": "string"
+				},
 				"started_at": {
 					"type": "string",
 					"format": "date-time"
@@ -11689,6 +11711,12 @@
 		"codersdk.AIBridgeSessionThreadsTokenUsage": {
 			"type": "object",
 			"properties": {
+				"cache_read_input_tokens": {
+					"type": "integer"
+				},
+				"cache_write_input_tokens": {
+					"type": "integer"
+				},
 				"input_tokens": {
 					"type": "integer"
 				},
@@ -11704,6 +11732,12 @@
 		"codersdk.AIBridgeSessionTokenUsageSummary": {
 			"type": "object",
 			"properties": {
+				"cache_read_input_tokens": {
+					"type": "integer"
+				},
+				"cache_write_input_tokens": {
+					"type": "integer"
+				},
 				"input_tokens": {
 					"type": "integer"
 				},
@@ -11750,6 +11784,12 @@
 		"codersdk.AIBridgeTokenUsage": {
 			"type": "object",
 			"properties": {
+				"cache_read_input_tokens": {
+					"type": "integer"
+				},
+				"cache_write_input_tokens": {
+					"type": "integer"
+				},
 				"created_at": {
 					"type": "string",
 					"format": "date-time"
@@ -12699,6 +12739,9 @@
 				},
 				"count": {
 					"type": "integer"
+				},
+				"count_cap": {
+					"type": "integer"
 				}
 			}
 		},
@@ -12999,6 +13042,9 @@
 				},
 				"count": {
 					"type": "integer"
+				},
+				"count_cap": {
+					"type": "integer"
 				}
 			}
 		},
@@ -13083,6 +13129,17 @@
 				}
 			}
 		},
+		"codersdk.CreateFirstUserOnboardingInfo": {
+			"type": "object",
+			"properties": {
+				"newsletter_marketing": {
+					"type": "boolean"
+				},
+				"newsletter_releases": {
+					"type": "boolean"
+				}
+			}
+		},
 		"codersdk.CreateFirstUserRequest": {
 			"type": "object",
 			"required": ["email", "password", "username"],
@@ -13093,6 +13150,9 @@
 				"name": {
 					"type": "string"
 				},
+				"onboarding_info": {
+					"$ref": "#/definitions/codersdk.CreateFirstUserOnboardingInfo"
+				},
 				"password": {
 					"type": "string"
 				},

coderd/apikey.go

@@ -582,5 +582,20 @@ func (api *API) createAPIKey(ctx context.Context, params apikey.CreateParams) (*
 		Value:    sessionToken,
 		Path:     "/",
 		HttpOnly: true,
+		// MaxAge is set so the browser persists the cookie to disk rather
+		// than keeping it in memory as a session cookie. Standalone PWAs
+		// (display: standalone) run in their own browser process, and
+		// mobile OSes kill that process when the app is swiped away —
+		// deleting in-memory cookies and forcing an unexpected login.
+		//
+		// We use a long static value (1 year) instead of the key's
+		// LifetimeSeconds because the server refreshes the key's
+		// ExpiresAt on activity but does not re-set the cookie. Tying
+		// MaxAge to the key lifetime would cause the cookie to expire
+		// client-side even when the server-side key is still valid.
+		//
+		// Security is not affected: the server validates ExpiresAt on
+		// every request regardless of the cookie's MaxAge.
+		MaxAge: int((365 * 24 * time.Hour).Seconds()),
 	}), &newkey, nil
 }

coderd/apikey_test.go

@@ -394,6 +394,55 @@ func TestSessionExpiry(t *testing.T) {
 	}
 }
 
+// TestSessionCookieMaxAge verifies that the session cookie is a persistent
+// cookie (has MaxAge set) rather than a session cookie. Standalone PWAs
+// run in their own browser process and mobile OSes purge in-memory
+// (session) cookies when that process is killed, so the cookie must be
+// persisted to disk.
+func TestSessionCookieMaxAge(t *testing.T) {
+	t.Parallel()
+
+	ctx, cancel := context.WithTimeout(context.Background(), testutil.WaitLong)
+	defer cancel()
+
+	client := coderdtest.New(t, nil)
+
+	// Create the first user (password-based login).
+	req := codersdk.CreateFirstUserRequest{
+		Email:    "testuser@coder.com",
+		Username: "testuser",
+		Password: "SomeSecurePassword!",
+	}
+	_, err := client.CreateFirstUser(ctx, req)
+	require.NoError(t, err)
+
+	// Login via the raw HTTP endpoint so we can inspect the Set-Cookie header.
+	loginURL, err := client.URL.Parse("/api/v2/users/login")
+	require.NoError(t, err)
+
+	res, err := client.Request(ctx, http.MethodPost, loginURL.String(), codersdk.LoginWithPasswordRequest{
+		Email:    req.Email,
+		Password: req.Password,
+	})
+	require.NoError(t, err)
+	defer res.Body.Close()
+	require.Equal(t, http.StatusCreated, res.StatusCode)
+
+	oneYear := int((365 * 24 * time.Hour).Seconds())
+	var found bool
+	for _, cookie := range res.Cookies() {
+		if cookie.Name == codersdk.SessionTokenCookie {
+			// MaxAge should be set to a long value so the browser
+			// persists the cookie to disk. The server handles real
+			// expiry via the API key's ExpiresAt field.
+			require.Equal(t, oneYear, cookie.MaxAge,
+				"Session cookie MaxAge should be set to 1 year for disk persistence")
+			found = true
+		}
+	}
+	require.True(t, found, "session cookie should be present in login response")
+}
+
 func TestAPIKey_OK(t *testing.T) {
 	t.Parallel()
 

coderd/audit.go

@@ -26,6 +26,11 @@ import (
 	"github.com/coder/coder/v2/codersdk"
 )
 
+// Limit the count query to avoid a slow sequential scan due to joins
+// on a large table. Set to 0 to disable capping (but also see the note
+// in the SQL query).
+const auditLogCountCap = 2000
+
 // @Summary Get audit logs
 // @ID get-audit-logs
 // @Security CoderSessionToken
@@ -66,7 +71,7 @@ func (api *API) auditLogs(rw http.ResponseWriter, r *http.Request) {
 		countFilter.Username = ""
 	}
 
-	// Use the same filters to count the number of audit logs
+	countFilter.CountCap = auditLogCountCap
 	count, err := api.Database.CountAuditLogs(ctx, countFilter)
 	if dbauthz.IsNotAuthorizedError(err) {
 		httpapi.Forbidden(rw)
@@ -81,6 +86,7 @@ func (api *API) auditLogs(rw http.ResponseWriter, r *http.Request) {
 		httpapi.Write(ctx, rw, http.StatusOK, codersdk.AuditLogResponse{
 			AuditLogs: []codersdk.AuditLog{},
 			Count:     0,
+			CountCap:  auditLogCountCap,
 		})
 		return
 	}
@@ -98,6 +104,7 @@ func (api *API) auditLogs(rw http.ResponseWriter, r *http.Request) {
 	httpapi.Write(ctx, rw, http.StatusOK, codersdk.AuditLogResponse{
 		AuditLogs: api.convertAuditLogs(ctx, dblogs),
 		Count:     count,
+		CountCap:  auditLogCountCap,
 	})
 }
 

coderd/coderd.go

@@ -168,6 +168,7 @@ type Options struct {
 	ConnectionLogger               connectionlog.ConnectionLogger
 	AgentConnectionUpdateFrequency time.Duration
 	AgentInactiveDisconnectTimeout time.Duration
+	ChatdInstructionLookupTimeout  time.Duration
 	AWSCertificates                awsidentity.Certificates
 	Authorizer                     rbac.Authorizer
 	AzureCertificates              x509.VerifyOptions
@@ -782,9 +783,10 @@ func New(options *Options) *API {
 			ReplicaID:                      api.ID,
 			SubscribeFn:                    options.ChatSubscribeFn,
 			MaxChatsPerAcquire:             int32(maxChatsPerAcquire), //nolint:gosec // maxChatsPerAcquire is clamped to int32 range above.
-			ProviderAPIKeys:                chatProviderAPIKeysFromDeploymentValues(options.DeploymentValues),
+			ProviderAPIKeys:                ChatProviderAPIKeysFromDeploymentValues(options.DeploymentValues),
 			AgentConn:                      api.agentProvider.AgentConn,
 			AgentInactiveDisconnectTimeout: api.AgentInactiveDisconnectTimeout,
+			InstructionLookupTimeout:       options.ChatdInstructionLookupTimeout,
 			CreateWorkspace:                api.chatCreateWorkspace,
 			StartWorkspace:                 api.chatStartWorkspace,
 			Pubsub:                         options.Pubsub,
@@ -1221,6 +1223,13 @@ func New(options *Options) *API {
 					r.Delete("/", api.deleteChatUsageLimitGroupOverride)
 				})
 			})
+			r.Route("/user-provider-configs", func(r chi.Router) {
+				r.Get("/", api.listUserChatProviderConfigs)
+				r.Route("/{providerConfig}", func(r chi.Router) {
+					r.Put("/", api.upsertUserChatProviderKey)
+					r.Delete("/", api.deleteUserChatProviderKey)
+				})
+			})
 			r.Route("/{chat}", func(r chi.Router) {
 				r.Use(httpmw.ExtractChatParam(options.Database))
 				r.Get("/", api.getChat)

coderd/coderdtest/coderdtest.go

@@ -149,12 +149,13 @@ type Options struct {
 	OneTimePasscodeValidityPeriod time.Duration
 
 	// IncludeProvisionerDaemon when true means to start an in-memory provisionerD
-	IncludeProvisionerDaemon    bool
-	ProvisionerDaemonVersion    string
-	ProvisionerDaemonTags       map[string]string
-	MetricsCacheRefreshInterval time.Duration
-	AgentStatsRefreshInterval   time.Duration
-	DeploymentValues            *codersdk.DeploymentValues
+	IncludeProvisionerDaemon      bool
+	ChatdInstructionLookupTimeout time.Duration
+	ProvisionerDaemonVersion      string
+	ProvisionerDaemonTags         map[string]string
+	MetricsCacheRefreshInterval   time.Duration
+	AgentStatsRefreshInterval     time.Duration
+	DeploymentValues              *codersdk.DeploymentValues
 
 	// Set update check options to enable update check.
 	UpdateCheckOptions *updatecheck.Options
@@ -575,6 +576,7 @@ func NewOptions(t testing.TB, options *Options) (func(http.Handler), context.Can
 			// Force a long disconnection timeout to ensure
 			// agents are not marked as disconnected during slow tests.
 			AgentInactiveDisconnectTimeout: testutil.WaitShort,
+			ChatdInstructionLookupTimeout:  options.ChatdInstructionLookupTimeout,
 			AccessURL:                      accessURL,
 			AppHostname:                    options.AppHostname,
 			AppHostnameRegex:               appHostnameRegex,

coderd/connectionlog/connectionlog.go

@@ -90,8 +90,8 @@ func (m *FakeConnectionLogger) Contains(t testing.TB, expected database.UpsertCo
 			t.Logf("connection log %d: expected Code %d, got %d", idx+1, expected.Code.Int32, cl.Code.Int32)
 			continue
 		}
-		if expected.Ip.Valid && cl.Ip.IPNet.String() != expected.Ip.IPNet.String() {
-			t.Logf("connection log %d: expected IP %s, got %s", idx+1, expected.Ip.IPNet, cl.Ip.IPNet)
+		if expected.IP.Valid && cl.IP.IPNet.String() != expected.IP.IPNet.String() {
+			t.Logf("connection log %d: expected IP %s, got %s", idx+1, expected.IP.IPNet, cl.IP.IPNet)
 			continue
 		}
 		if expected.UserAgent.Valid && cl.UserAgent.String != expected.UserAgent.String {

coderd/database/check_constraint.go

@@ -10,6 +10,7 @@ const (
 	CheckChatModelConfigsCompressionThresholdCheck   CheckConstraint = "chat_model_configs_compression_threshold_check"     // chat_model_configs
 	CheckChatModelConfigsContextLimitCheck           CheckConstraint = "chat_model_configs_context_limit_check"             // chat_model_configs
 	CheckChatProvidersProviderCheck                  CheckConstraint = "chat_providers_provider_check"                      // chat_providers
+	CheckValidCredentialPolicy                       CheckConstraint = "valid_credential_policy"                            // chat_providers
 	CheckChatUsageLimitConfigDefaultLimitMicrosCheck CheckConstraint = "chat_usage_limit_config_default_limit_micros_check" // chat_usage_limit_config
 	CheckChatUsageLimitConfigPeriodCheck             CheckConstraint = "chat_usage_limit_config_period_check"               // chat_usage_limit_config
 	CheckChatUsageLimitConfigSingletonCheck          CheckConstraint = "chat_usage_limit_config_singleton_check"            // chat_usage_limit_config
@@ -32,4 +33,5 @@ const (
 	CheckTelemetryLockEventTypeConstraint            CheckConstraint = "telemetry_lock_event_type_constraint"               // telemetry_locks
 	CheckValidationMonotonicOrder                    CheckConstraint = "validation_monotonic_order"                         // template_version_parameters
 	CheckUsageEventTypeCheck                         CheckConstraint = "usage_event_type_check"                             // usage_events
+	CheckUserChatProviderKeysAPIKeyCheck             CheckConstraint = "user_chat_provider_keys_api_key_check"              // user_chat_provider_keys
 )

coderd/database/db2sdk/db2sdk.go

@@ -999,15 +999,16 @@ func AIBridgeInterception(interception database.AIBridgeInterception, initiator
 		return sdkToolUsages[i].CreatedAt.Before(sdkToolUsages[j].CreatedAt)
 	})
 	intc := codersdk.AIBridgeInterception{
-		ID:          interception.ID,
-		Initiator:   MinimalUserFromVisibleUser(initiator),
-		Provider:    interception.Provider,
-		Model:       interception.Model,
-		Metadata:    jsonOrEmptyMap(interception.Metadata),
-		StartedAt:   interception.StartedAt,
-		TokenUsages: sdkTokenUsages,
-		UserPrompts: sdkUserPrompts,
-		ToolUsages:  sdkToolUsages,
+		ID:           interception.ID,
+		Initiator:    MinimalUserFromVisibleUser(initiator),
+		Provider:     interception.Provider,
+		ProviderName: interception.ProviderName,
+		Model:        interception.Model,
+		Metadata:     jsonOrEmptyMap(interception.Metadata),
+		StartedAt:    interception.StartedAt,
+		TokenUsages:  sdkTokenUsages,
+		UserPrompts:  sdkUserPrompts,
+		ToolUsages:   sdkToolUsages,
 	}
 	if interception.APIKeyID.Valid {
 		intc.APIKeyID = &interception.APIKeyID.String
@@ -1036,8 +1037,10 @@ func AIBridgeSession(row database.ListAIBridgeSessionsRow) codersdk.AIBridgeSess
 		StartedAt: row.StartedAt,
 		Threads:   row.Threads,
 		TokenUsageSummary: codersdk.AIBridgeSessionTokenUsageSummary{
-			InputTokens:  row.InputTokens,
-			OutputTokens: row.OutputTokens,
+			InputTokens:           row.InputTokens,
+			OutputTokens:          row.OutputTokens,
+			CacheReadInputTokens:  row.CacheReadInputTokens,
+			CacheWriteInputTokens: row.CacheWriteInputTokens,
 		},
 	}
 	// Ensure non-nil slices for JSON serialization.
@@ -1061,13 +1064,15 @@ func AIBridgeSession(row database.ListAIBridgeSessionsRow) codersdk.AIBridgeSess
 
 func AIBridgeTokenUsage(usage database.AIBridgeTokenUsage) codersdk.AIBridgeTokenUsage {
 	return codersdk.AIBridgeTokenUsage{
-		ID:                 usage.ID,
-		InterceptionID:     usage.InterceptionID,
-		ProviderResponseID: usage.ProviderResponseID,
-		InputTokens:        usage.InputTokens,
-		OutputTokens:       usage.OutputTokens,
-		Metadata:           jsonOrEmptyMap(usage.Metadata),
-		CreatedAt:          usage.CreatedAt,
+		ID:                    usage.ID,
+		InterceptionID:        usage.InterceptionID,
+		ProviderResponseID:    usage.ProviderResponseID,
+		InputTokens:           usage.InputTokens,
+		OutputTokens:          usage.OutputTokens,
+		CacheReadInputTokens:  usage.CacheReadInputTokens,
+		CacheWriteInputTokens: usage.CacheWriteInputTokens,
+		Metadata:              jsonOrEmptyMap(usage.Metadata),
+		CreatedAt:             usage.CreatedAt,
 	}
 }
 
@@ -1178,9 +1183,11 @@ func AIBridgeSessionThreads(
 		PageStartedAt: pageStartedAt,
 		PageEndedAt:   pageEndedAt,
 		TokenUsageSummary: codersdk.AIBridgeSessionThreadsTokenUsage{
-			InputTokens:  session.InputTokens,
-			OutputTokens: session.OutputTokens,
-			Metadata:     sessionTokenMeta,
+			InputTokens:           session.InputTokens,
+			OutputTokens:          session.OutputTokens,
+			CacheReadInputTokens:  session.CacheReadInputTokens,
+			CacheWriteInputTokens: session.CacheWriteInputTokens,
+			Metadata:              sessionTokenMeta,
 		},
 		Threads: threads,
 	}
@@ -1313,17 +1320,19 @@ func buildAIBridgeThread(
 
 // aggregateTokenUsage sums token usage rows and aggregates metadata.
 func aggregateTokenUsage(tokens []database.AIBridgeTokenUsage) codersdk.AIBridgeSessionThreadsTokenUsage {
-	var inputTokens, outputTokens int64
+	var inputTokens, outputTokens, cacheRead, cacheWrite int64
 	for _, tu := range tokens {
 		inputTokens += tu.InputTokens
 		outputTokens += tu.OutputTokens
-		// TODO: once https://github.com/coder/aibridge/issues/150 lands we
-		// should aggregate the other token types.
+		cacheRead += tu.CacheReadInputTokens
+		cacheWrite += tu.CacheWriteInputTokens
 	}
 	return codersdk.AIBridgeSessionThreadsTokenUsage{
-		InputTokens:  inputTokens,
-		OutputTokens: outputTokens,
-		Metadata:     aggregateTokenMetadata(tokens),
+		InputTokens:           inputTokens,
+		OutputTokens:          outputTokens,
+		CacheReadInputTokens:  cacheRead,
+		CacheWriteInputTokens: cacheWrite,
+		Metadata:              aggregateTokenMetadata(tokens),
 	}
 }
 
@@ -1519,7 +1528,10 @@ func nullInt64Ptr(v sql.NullInt64) *int64 {
 // Chat converts a database.Chat to a codersdk.Chat. It coalesces
 // nil slices and maps to empty values for JSON serialization and
 // derives RootChatID from the parent chain when not explicitly set.
-func Chat(c database.Chat, diffStatus *database.ChatDiffStatus) codersdk.Chat {
+// When diffStatus is non-nil the response includes diff metadata.
+// When files is non-empty the response includes file metadata;
+// pass nil to omit the files field (e.g. list endpoints).
+func Chat(c database.Chat, diffStatus *database.ChatDiffStatus, files []database.GetChatFileMetadataByChatIDRow) codersdk.Chat {
 	mcpServerIDs := c.MCPServerIDs
 	if mcpServerIDs == nil {
 		mcpServerIDs = []uuid.UUID{}
@@ -1572,6 +1584,19 @@ func Chat(c database.Chat, diffStatus *database.ChatDiffStatus) codersdk.Chat {
 		convertedDiffStatus := ChatDiffStatus(c.ID, diffStatus)
 		chat.DiffStatus = &convertedDiffStatus
 	}
+	if len(files) > 0 {
+		chat.Files = make([]codersdk.ChatFileMetadata, 0, len(files))
+		for _, row := range files {
+			chat.Files = append(chat.Files, codersdk.ChatFileMetadata{
+				ID:             row.ID,
+				OwnerID:        row.OwnerID,
+				OrganizationID: row.OrganizationID,
+				Name:           row.Name,
+				MimeType:       row.Mimetype,
+				CreatedAt:      row.CreatedAt,
+			})
+		}
+	}
 	if c.LastInjectedContext.Valid {
 		var parts []codersdk.ChatMessagePart
 		// Internal fields are stripped at write time in
@@ -1595,9 +1620,9 @@ func ChatRows(rows []database.GetChatsRow, diffStatusesByChatID map[uuid.UUID]da
 	for i, row := range rows {
 		diffStatus, ok := diffStatusesByChatID[row.Chat.ID]
 		if ok {
-			result[i] = Chat(row.Chat, &diffStatus)
+			result[i] = Chat(row.Chat, &diffStatus, nil)
 		} else {
-			result[i] = Chat(row.Chat, nil)
+			result[i] = Chat(row.Chat, nil, nil)
 			if diffStatusesByChatID != nil {
 				emptyDiffStatus := ChatDiffStatus(row.Chat.ID, nil)
 				result[i].DiffStatus = &emptyDiffStatus

coderd/database/db2sdk/db2sdk_test.go

@@ -259,11 +259,13 @@ func TestAIBridgeInterception(t *testing.T) {
 			},
 			tokenUsages: []database.AIBridgeTokenUsage{
 				{
-					ID:                 uuid.New(),
-					InterceptionID:     interceptionID,
-					ProviderResponseID: "resp-123",
-					InputTokens:        100,
-					OutputTokens:       200,
+					ID:                    uuid.New(),
+					InterceptionID:        interceptionID,
+					ProviderResponseID:    "resp-123",
+					InputTokens:           100,
+					OutputTokens:          200,
+					CacheReadInputTokens:  50,
+					CacheWriteInputTokens: 10,
 					Metadata: pqtype.NullRawMessage{
 						RawMessage: json.RawMessage(`{"cache":"hit"}`),
 						Valid:      true,
@@ -413,6 +415,8 @@ func TestAIBridgeInterception(t *testing.T) {
 				require.Equal(t, tu.ProviderResponseID, result.TokenUsages[i].ProviderResponseID)
 				require.Equal(t, tu.InputTokens, result.TokenUsages[i].InputTokens)
 				require.Equal(t, tu.OutputTokens, result.TokenUsages[i].OutputTokens)
+				require.Equal(t, tu.CacheReadInputTokens, result.TokenUsages[i].CacheReadInputTokens)
+				require.Equal(t, tu.CacheWriteInputTokens, result.TokenUsages[i].CacheWriteInputTokens)
 			}
 
 			// Verify user prompts are converted correctly.
@@ -557,14 +561,26 @@ func TestChat_AllFieldsPopulated(t *testing.T) {
 		ChatID: input.ID,
 	}
 
-	got := db2sdk.Chat(input, diffStatus)
+	fileRows := []database.GetChatFileMetadataByChatIDRow{
+		{
+			ID:             uuid.New(),
+			OwnerID:        input.OwnerID,
+			OrganizationID: uuid.New(),
+			Name:           "test.png",
+			Mimetype:       "image/png",
+			CreatedAt:      now,
+		},
+	}
+
+	got := db2sdk.Chat(input, diffStatus, fileRows)
 
 	v := reflect.ValueOf(got)
 	typ := v.Type()
 	// HasUnread is populated by ChatRows (which joins the
-	// read-cursor query), not by Chat, so it is expected
-	// to remain zero here.
-	skip := map[string]bool{"HasUnread": true}
+	// read-cursor query), not by Chat. Warnings is a transient
+	// field populated by handlers, not the converter. Both are
+	// expected to remain zero here.
+	skip := map[string]bool{"HasUnread": true, "Warnings": true}
 	for i := range typ.NumField() {
 		field := typ.Field(i)
 		if skip[field.Name] {
@@ -577,6 +593,112 @@ func TestChat_AllFieldsPopulated(t *testing.T) {
 	}
 }
 
+func TestChat_FileMetadataConversion(t *testing.T) {
+	t.Parallel()
+
+	ownerID := uuid.New()
+	orgID := uuid.New()
+	fileID := uuid.New()
+	now := dbtime.Now()
+
+	chat := database.Chat{
+		ID:                uuid.New(),
+		OwnerID:           ownerID,
+		LastModelConfigID: uuid.New(),
+		Title:             "file metadata test",
+		Status:            database.ChatStatusWaiting,
+		CreatedAt:         now,
+		UpdatedAt:         now,
+	}
+
+	rows := []database.GetChatFileMetadataByChatIDRow{
+		{
+			ID:             fileID,
+			OwnerID:        ownerID,
+			OrganizationID: orgID,
+			Name:           "screenshot.png",
+			Mimetype:       "image/png",
+			CreatedAt:      now,
+		},
+	}
+
+	result := db2sdk.Chat(chat, nil, rows)
+
+	require.Len(t, result.Files, 1)
+	f := result.Files[0]
+	require.Equal(t, fileID, f.ID)
+	require.Equal(t, ownerID, f.OwnerID, "OwnerID must be mapped from DB row")
+	require.Equal(t, orgID, f.OrganizationID, "OrganizationID must be mapped from DB row")
+	require.Equal(t, "screenshot.png", f.Name)
+	require.Equal(t, "image/png", f.MimeType)
+	require.Equal(t, now, f.CreatedAt)
+
+	// Verify JSON serialization uses snake_case for mime_type.
+	data, err := json.Marshal(f)
+	require.NoError(t, err)
+	require.Contains(t, string(data), `"mime_type"`)
+	require.NotContains(t, string(data), `"mimetype"`)
+}
+
+func TestChat_NilFilesOmitted(t *testing.T) {
+	t.Parallel()
+
+	chat := database.Chat{
+		ID:                uuid.New(),
+		OwnerID:           uuid.New(),
+		LastModelConfigID: uuid.New(),
+		Title:             "no files",
+		Status:            database.ChatStatusWaiting,
+		CreatedAt:         dbtime.Now(),
+		UpdatedAt:         dbtime.Now(),
+	}
+
+	result := db2sdk.Chat(chat, nil, nil)
+	require.Empty(t, result.Files)
+}
+
+func TestChat_MultipleFiles(t *testing.T) {
+	t.Parallel()
+
+	now := dbtime.Now()
+	file1 := uuid.New()
+	file2 := uuid.New()
+
+	chat := database.Chat{
+		ID:                uuid.New(),
+		OwnerID:           uuid.New(),
+		LastModelConfigID: uuid.New(),
+		Title:             "multi file test",
+		Status:            database.ChatStatusWaiting,
+		CreatedAt:         now,
+		UpdatedAt:         now,
+	}
+
+	rows := []database.GetChatFileMetadataByChatIDRow{
+		{
+			ID:             file1,
+			OwnerID:        chat.OwnerID,
+			OrganizationID: uuid.New(),
+			Name:           "a.png",
+			Mimetype:       "image/png",
+			CreatedAt:      now,
+		},
+		{
+			ID:             file2,
+			OwnerID:        chat.OwnerID,
+			OrganizationID: uuid.New(),
+			Name:           "b.txt",
+			Mimetype:       "text/plain",
+			CreatedAt:      now,
+		},
+	}
+
+	result := db2sdk.Chat(chat, nil, rows)
+	require.Len(t, result.Files, 2)
+	require.Equal(t, "a.png", result.Files[0].Name)
+	require.Equal(t, "b.txt", result.Files[1].Name)
+}
+
 func TestChatQueuedMessage_MalformedContent(t *testing.T) {
 	t.Parallel()
 

coderd/database/dbauthz/dbauthz.go

@@ -1570,13 +1570,13 @@ func (q *querier) AllUserIDs(ctx context.Context, includeSystem bool) ([]uuid.UU
 	return q.db.AllUserIDs(ctx, includeSystem)
 }
 
-func (q *querier) ArchiveChatByID(ctx context.Context, id uuid.UUID) error {
+func (q *querier) ArchiveChatByID(ctx context.Context, id uuid.UUID) ([]database.Chat, error) {
 	chat, err := q.db.GetChatByID(ctx, id)
 	if err != nil {
-		return err
+		return nil, err
 	}
 	if err := q.authorizeContext(ctx, policy.ActionUpdate, chat); err != nil {
-		return err
+		return nil, err
 	}
 	return q.db.ArchiveChatByID(ctx, id)
 }
@@ -1627,6 +1627,13 @@ func (q *querier) BatchUpdateWorkspaceNextStartAt(ctx context.Context, arg datab
 	return q.db.BatchUpdateWorkspaceNextStartAt(ctx, arg)
 }
 
+func (q *querier) BatchUpsertConnectionLogs(ctx context.Context, arg database.BatchUpsertConnectionLogsParams) error {
+	if err := q.authorizeContext(ctx, policy.ActionUpdate, rbac.ResourceConnectionLog); err != nil {
+		return err
+	}
+	return q.db.BatchUpsertConnectionLogs(ctx, arg)
+}
+
 func (q *querier) BulkMarkNotificationMessagesFailed(ctx context.Context, arg database.BulkMarkNotificationMessagesFailedParams) (int64, error) {
 	if err := q.authorizeContext(ctx, policy.ActionUpdate, rbac.ResourceNotificationMessage); err != nil {
 		return 0, err
@@ -2137,6 +2144,17 @@ func (q *querier) DeleteUserChatCompactionThreshold(ctx context.Context, arg dat
 	return q.db.DeleteUserChatCompactionThreshold(ctx, arg)
 }
 
+func (q *querier) DeleteUserChatProviderKey(ctx context.Context, arg database.DeleteUserChatProviderKeyParams) error {
+	u, err := q.db.GetUserByID(ctx, arg.UserID)
+	if err != nil {
+		return err
+	}
+	if err := q.authorizeContext(ctx, policy.ActionUpdatePersonal, u); err != nil {
+		return err
+	}
+	return q.db.DeleteUserChatProviderKey(ctx, arg)
+}
+
 func (q *querier) DeleteUserSecret(ctx context.Context, id uuid.UUID) error {
 	// First get the secret to check ownership
 	secret, err := q.GetUserSecret(ctx, id)
@@ -2565,6 +2583,10 @@ func (q *querier) GetChatFileByID(ctx context.Context, id uuid.UUID) (database.C
 	return file, nil
 }
 
+func (q *querier) GetChatFileMetadataByChatID(ctx context.Context, chatID uuid.UUID) ([]database.GetChatFileMetadataByChatIDRow, error) {
+	return fetchWithPostFilter(q.auth, policy.ActionRead, q.db.GetChatFileMetadataByChatID)(ctx, chatID)
+}
+
 func (q *querier) GetChatFilesByIDs(ctx context.Context, ids []uuid.UUID) ([]database.ChatFile, error) {
 	files, err := q.db.GetChatFilesByIDs(ctx, ids)
 	if err != nil {
@@ -3635,18 +3657,18 @@ func (q *querier) GetTailnetPeers(ctx context.Context, id uuid.UUID) ([]database
 	return q.db.GetTailnetPeers(ctx, id)
 }
 
-func (q *querier) GetTailnetTunnelPeerBindings(ctx context.Context, srcID uuid.UUID) ([]database.GetTailnetTunnelPeerBindingsRow, error) {
+func (q *querier) GetTailnetTunnelPeerBindingsBatch(ctx context.Context, ids []uuid.UUID) ([]database.GetTailnetTunnelPeerBindingsBatchRow, error) {
 	if err := q.authorizeContext(ctx, policy.ActionRead, rbac.ResourceTailnetCoordinator); err != nil {
 		return nil, err
 	}
-	return q.db.GetTailnetTunnelPeerBindings(ctx, srcID)
+	return q.db.GetTailnetTunnelPeerBindingsBatch(ctx, ids)
 }
 
-func (q *querier) GetTailnetTunnelPeerIDs(ctx context.Context, srcID uuid.UUID) ([]database.GetTailnetTunnelPeerIDsRow, error) {
+func (q *querier) GetTailnetTunnelPeerIDsBatch(ctx context.Context, ids []uuid.UUID) ([]database.GetTailnetTunnelPeerIDsBatchRow, error) {
 	if err := q.authorizeContext(ctx, policy.ActionRead, rbac.ResourceTailnetCoordinator); err != nil {
 		return nil, err
 	}
-	return q.db.GetTailnetTunnelPeerIDs(ctx, srcID)
+	return q.db.GetTailnetTunnelPeerIDsBatch(ctx, ids)
 }
 
 func (q *querier) GetTaskByID(ctx context.Context, id uuid.UUID) (database.Task, error) {
@@ -4024,6 +4046,17 @@ func (q *querier) GetUserChatCustomPrompt(ctx context.Context, userID uuid.UUID)
 	return q.db.GetUserChatCustomPrompt(ctx, userID)
 }
 
+func (q *querier) GetUserChatProviderKeys(ctx context.Context, userID uuid.UUID) ([]database.UserChatProviderKey, error) {
+	u, err := q.db.GetUserByID(ctx, userID)
+	if err != nil {
+		return nil, err
+	}
+	if err := q.authorizeContext(ctx, policy.ActionReadPersonal, u); err != nil {
+		return nil, err
+	}
+	return q.db.GetUserChatProviderKeys(ctx, userID)
+}
+
 func (q *querier) GetUserChatSpendInPeriod(ctx context.Context, arg database.GetUserChatSpendInPeriodParams) (int64, error) {
 	if err := q.authorizeContext(ctx, policy.ActionRead, rbac.ResourceChat.WithOwner(arg.UserID.String())); err != nil {
 		return 0, err
@@ -5364,6 +5397,17 @@ func (q *querier) InsertWorkspaceResourceMetadata(ctx context.Context, arg datab
 	return q.db.InsertWorkspaceResourceMetadata(ctx, arg)
 }
 
+func (q *querier) LinkChatFiles(ctx context.Context, arg database.LinkChatFilesParams) (int32, error) {
+	chat, err := q.db.GetChatByID(ctx, arg.ChatID)
+	if err != nil {
+		return 0, err
+	}
+	if err := q.authorizeContext(ctx, policy.ActionUpdate, chat); err != nil {
+		return 0, err
+	}
+	return q.db.LinkChatFiles(ctx, arg)
+}
+
 func (q *querier) ListAIBridgeClients(ctx context.Context, arg database.ListAIBridgeClientsParams) ([]string, error) {
 	prep, err := prepareSQLFilter(ctx, q.auth, policy.ActionRead, rbac.ResourceAibridgeInterception.Type)
 	if err != nil {
@@ -5649,13 +5693,13 @@ func (q *querier) TryAcquireLock(ctx context.Context, id int64) (bool, error) {
 	return q.db.TryAcquireLock(ctx, id)
 }
 
-func (q *querier) UnarchiveChatByID(ctx context.Context, id uuid.UUID) error {
+func (q *querier) UnarchiveChatByID(ctx context.Context, id uuid.UUID) ([]database.Chat, error) {
 	chat, err := q.db.GetChatByID(ctx, id)
 	if err != nil {
-		return err
+		return nil, err
 	}
 	if err := q.authorizeContext(ctx, policy.ActionUpdate, chat); err != nil {
-		return err
+		return nil, err
 	}
 	return q.db.UnarchiveChatByID(ctx, id)
 }
@@ -5738,15 +5782,15 @@ func (q *querier) UpdateChatByID(ctx context.Context, arg database.UpdateChatByI
 	return q.db.UpdateChatByID(ctx, arg)
 }
 
-func (q *querier) UpdateChatHeartbeat(ctx context.Context, arg database.UpdateChatHeartbeatParams) (int64, error) {
-	chat, err := q.db.GetChatByID(ctx, arg.ID)
-	if err != nil {
-		return 0, err
+func (q *querier) UpdateChatHeartbeats(ctx context.Context, arg database.UpdateChatHeartbeatsParams) ([]uuid.UUID, error) {
+	// The batch heartbeat is a system-level operation filtered by
+	// worker_id. Authorization is enforced by the AsChatd context
+	// at the call site rather than per-row, because checking each
+	// row individually would defeat the purpose of batching.
+	if err := q.authorizeContext(ctx, policy.ActionUpdate, rbac.ResourceChat); err != nil {
+		return nil, err
 	}
-	if err := q.authorizeContext(ctx, policy.ActionUpdate, chat); err != nil {
-		return 0, err
-	}
-	return q.db.UpdateChatHeartbeat(ctx, arg)
+	return q.db.UpdateChatHeartbeats(ctx, arg)
 }
 
 func (q *querier) UpdateChatLabelsByID(ctx context.Context, arg database.UpdateChatLabelsByIDParams) (database.Chat, error) {
@@ -6454,6 +6498,17 @@ func (q *querier) UpdateUserChatCustomPrompt(ctx context.Context, arg database.U
 	return q.db.UpdateUserChatCustomPrompt(ctx, arg)
 }
 
+func (q *querier) UpdateUserChatProviderKey(ctx context.Context, arg database.UpdateUserChatProviderKeyParams) (database.UserChatProviderKey, error) {
+	u, err := q.db.GetUserByID(ctx, arg.UserID)
+	if err != nil {
+		return database.UserChatProviderKey{}, err
+	}
+	if err := q.authorizeContext(ctx, policy.ActionUpdatePersonal, u); err != nil {
+		return database.UserChatProviderKey{}, err
+	}
+	return q.db.UpdateUserChatProviderKey(ctx, arg)
+}
+
 func (q *querier) UpdateUserDeletedByID(ctx context.Context, id uuid.UUID) error {
 	return deleteQ(q.log, q.auth, q.db.GetUserByID, q.db.UpdateUserDeletedByID)(ctx, id)
 }
@@ -7032,13 +7087,6 @@ func (q *querier) UpsertChatWorkspaceTTL(ctx context.Context, workspaceTtl strin
 	return q.db.UpsertChatWorkspaceTTL(ctx, workspaceTtl)
 }
 
-func (q *querier) UpsertConnectionLog(ctx context.Context, arg database.UpsertConnectionLogParams) (database.ConnectionLog, error) {
-	if err := q.authorizeContext(ctx, policy.ActionUpdate, rbac.ResourceConnectionLog); err != nil {
-		return database.ConnectionLog{}, err
-	}
-	return q.db.UpsertConnectionLog(ctx, arg)
-}
-
 func (q *querier) UpsertDefaultProxy(ctx context.Context, arg database.UpsertDefaultProxyParams) error {
 	if err := q.authorizeContext(ctx, policy.ActionUpdate, rbac.ResourceSystem); err != nil {
 		return err
@@ -7181,6 +7229,17 @@ func (q *querier) UpsertTemplateUsageStats(ctx context.Context) error {
 	return q.db.UpsertTemplateUsageStats(ctx)
 }
 
+func (q *querier) UpsertUserChatProviderKey(ctx context.Context, arg database.UpsertUserChatProviderKeyParams) (database.UserChatProviderKey, error) {
+	u, err := q.db.GetUserByID(ctx, arg.UserID)
+	if err != nil {
+		return database.UserChatProviderKey{}, err
+	}
+	if err := q.authorizeContext(ctx, policy.ActionUpdatePersonal, u); err != nil {
+		return database.UserChatProviderKey{}, err
+	}
+	return q.db.UpsertUserChatProviderKey(ctx, arg)
+}
+
 func (q *querier) UpsertWebpushVAPIDKeys(ctx context.Context, arg database.UpsertWebpushVAPIDKeysParams) error {
 	if err := q.authorizeContext(ctx, policy.ActionUpdate, rbac.ResourceDeploymentConfig); err != nil {
 		return err

coderd/database/dbauthz/dbauthz_test.go

@@ -338,10 +338,9 @@ func (s *MethodTestSuite) TestAuditLogs() {
 }
 
 func (s *MethodTestSuite) TestConnectionLogs() {
-	s.Run("UpsertConnectionLog", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
-		ws := testutil.Fake(s.T(), faker, database.WorkspaceTable{})
-		arg := database.UpsertConnectionLogParams{Ip: defaultIPAddress(), Type: database.ConnectionTypeSsh, WorkspaceID: ws.ID, OrganizationID: ws.OrganizationID, ConnectionStatus: database.ConnectionStatusConnected, WorkspaceOwnerID: ws.OwnerID}
-		dbm.EXPECT().UpsertConnectionLog(gomock.Any(), arg).Return(database.ConnectionLog{}, nil).AnyTimes()
+	s.Run("BatchUpsertConnectionLogs", s.Mocked(func(dbm *dbmock.MockStore, _ *gofakeit.Faker, check *expects) {
+		arg := database.BatchUpsertConnectionLogsParams{}
+		dbm.EXPECT().BatchUpsertConnectionLogs(gomock.Any(), arg).Return(nil).AnyTimes()
 		check.Args(arg).Asserts(rbac.ResourceConnectionLog, policy.ActionUpdate)
 	}))
 	s.Run("GetConnectionLogsOffset", s.Mocked(func(dbm *dbmock.MockStore, _ *gofakeit.Faker, check *expects) {
@@ -392,14 +391,25 @@ func (s *MethodTestSuite) TestChats() {
 	s.Run("ArchiveChatByID", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
 		chat := testutil.Fake(s.T(), faker, database.Chat{})
 		dbm.EXPECT().GetChatByID(gomock.Any(), chat.ID).Return(chat, nil).AnyTimes()
-		dbm.EXPECT().ArchiveChatByID(gomock.Any(), chat.ID).Return(nil).AnyTimes()
-		check.Args(chat.ID).Asserts(chat, policy.ActionUpdate).Returns()
+		dbm.EXPECT().ArchiveChatByID(gomock.Any(), chat.ID).Return([]database.Chat{chat}, nil).AnyTimes()
+		check.Args(chat.ID).Asserts(chat, policy.ActionUpdate).Returns([]database.Chat{chat})
 	}))
 	s.Run("UnarchiveChatByID", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
 		chat := testutil.Fake(s.T(), faker, database.Chat{})
 		dbm.EXPECT().GetChatByID(gomock.Any(), chat.ID).Return(chat, nil).AnyTimes()
-		dbm.EXPECT().UnarchiveChatByID(gomock.Any(), chat.ID).Return(nil).AnyTimes()
-		check.Args(chat.ID).Asserts(chat, policy.ActionUpdate).Returns()
+		dbm.EXPECT().UnarchiveChatByID(gomock.Any(), chat.ID).Return([]database.Chat{chat}, nil).AnyTimes()
+		check.Args(chat.ID).Asserts(chat, policy.ActionUpdate).Returns([]database.Chat{chat})
+	}))
+	s.Run("LinkChatFiles", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
+		chat := testutil.Fake(s.T(), faker, database.Chat{})
+		arg := database.LinkChatFilesParams{
+			ChatID:       chat.ID,
+			MaxFileLinks: int32(codersdk.MaxChatFileIDs),
+			FileIds:      []uuid.UUID{uuid.New()},
+		}
+		dbm.EXPECT().GetChatByID(gomock.Any(), chat.ID).Return(chat, nil).AnyTimes()
+		dbm.EXPECT().LinkChatFiles(gomock.Any(), arg).Return(int32(0), nil).AnyTimes()
+		check.Args(arg).Asserts(chat, policy.ActionUpdate).Returns(int32(0))
 	}))
 	s.Run("PinChatByID", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
 		chat := testutil.Fake(s.T(), faker, database.Chat{})
@@ -577,6 +587,19 @@ func (s *MethodTestSuite) TestChats() {
 		dbm.EXPECT().GetChatFilesByIDs(gomock.Any(), []uuid.UUID{file.ID}).Return([]database.ChatFile{file}, nil).AnyTimes()
 		check.Args([]uuid.UUID{file.ID}).Asserts(rbac.ResourceChat.WithOwner(file.OwnerID.String()).InOrg(file.OrganizationID).WithID(file.ID), policy.ActionRead).Returns([]database.ChatFile{file})
 	}))
+	s.Run("GetChatFileMetadataByChatID", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
+		file := testutil.Fake(s.T(), faker, database.ChatFile{})
+		rows := []database.GetChatFileMetadataByChatIDRow{{
+			ID:             file.ID,
+			Name:           file.Name,
+			Mimetype:       file.Mimetype,
+			CreatedAt:      file.CreatedAt,
+			OwnerID:        file.OwnerID,
+			OrganizationID: file.OrganizationID,
+		}}
+		dbm.EXPECT().GetChatFileMetadataByChatID(gomock.Any(), file.ID).Return(rows, nil).AnyTimes()
+		check.Args(file.ID).Asserts(rbac.ResourceChat.WithOwner(file.OwnerID.String()).InOrg(file.OrganizationID).WithID(file.ID), policy.ActionRead).Returns(rows)
+	}))
 	s.Run("GetChatMessageByID", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
 		chat := testutil.Fake(s.T(), faker, database.Chat{})
 		msg := testutil.Fake(s.T(), faker, database.ChatMessage{ChatID: chat.ID})
@@ -721,7 +744,9 @@ func (s *MethodTestSuite) TestChats() {
 		check.Args(threshold).Asserts(rbac.ResourceChat, policy.ActionRead).Returns(chats)
 	}))
 	s.Run("InsertChat", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
-		arg := testutil.Fake(s.T(), faker, database.InsertChatParams{})
+		arg := testutil.Fake(s.T(), faker, database.InsertChatParams{
+			Status: database.ChatStatusWaiting,
+		})
 		chat := testutil.Fake(s.T(), faker, database.Chat{OwnerID: arg.OwnerID})
 		dbm.EXPECT().InsertChat(gomock.Any(), arg).Return(chat, nil).AnyTimes()
 		check.Args(arg).Asserts(rbac.ResourceChat.WithOwner(arg.OwnerID.String()), policy.ActionCreate).Returns(chat)
@@ -817,15 +842,15 @@ func (s *MethodTestSuite) TestChats() {
 		dbm.EXPECT().UpdateChatStatusPreserveUpdatedAt(gomock.Any(), arg).Return(chat, nil).AnyTimes()
 		check.Args(arg).Asserts(chat, policy.ActionUpdate).Returns(chat)
 	}))
-	s.Run("UpdateChatHeartbeat", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
-		chat := testutil.Fake(s.T(), faker, database.Chat{})
-		arg := database.UpdateChatHeartbeatParams{
-			ID:       chat.ID,
+	s.Run("UpdateChatHeartbeats", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
+		resultID := uuid.New()
+		arg := database.UpdateChatHeartbeatsParams{
+			IDs:      []uuid.UUID{resultID},
 			WorkerID: uuid.New(),
+			Now:      time.Now(),
 		}
-		dbm.EXPECT().GetChatByID(gomock.Any(), chat.ID).Return(chat, nil).AnyTimes()
-		dbm.EXPECT().UpdateChatHeartbeat(gomock.Any(), arg).Return(int64(1), nil).AnyTimes()
-		check.Args(arg).Asserts(chat, policy.ActionUpdate).Returns(int64(1))
+		dbm.EXPECT().UpdateChatHeartbeats(gomock.Any(), arg).Return([]uuid.UUID{resultID}, nil).AnyTimes()
+		check.Args(arg).Asserts(rbac.ResourceChat, policy.ActionUpdate).Returns([]uuid.UUID{resultID})
 	}))
 	s.Run("UpdateChatMessageByID", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
 		chat := testutil.Fake(s.T(), faker, database.Chat{})
@@ -2405,6 +2430,36 @@ func (s *MethodTestSuite) TestUser() {
 		dbm.EXPECT().GetUserChatCustomPrompt(gomock.Any(), u.ID).Return("my custom prompt", nil).AnyTimes()
 		check.Args(u.ID).Asserts(u, policy.ActionReadPersonal).Returns("my custom prompt")
 	}))
+	s.Run("GetUserChatProviderKeys", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
+		u := testutil.Fake(s.T(), faker, database.User{})
+		key := testutil.Fake(s.T(), faker, database.UserChatProviderKey{UserID: u.ID})
+		dbm.EXPECT().GetUserByID(gomock.Any(), u.ID).Return(u, nil).AnyTimes()
+		dbm.EXPECT().GetUserChatProviderKeys(gomock.Any(), u.ID).Return([]database.UserChatProviderKey{key}, nil).AnyTimes()
+		check.Args(u.ID).Asserts(u, policy.ActionReadPersonal).Returns([]database.UserChatProviderKey{key})
+	}))
+	s.Run("DeleteUserChatProviderKey", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
+		u := testutil.Fake(s.T(), faker, database.User{})
+		arg := database.DeleteUserChatProviderKeyParams{UserID: u.ID, ChatProviderID: uuid.New()}
+		dbm.EXPECT().GetUserByID(gomock.Any(), u.ID).Return(u, nil).AnyTimes()
+		dbm.EXPECT().DeleteUserChatProviderKey(gomock.Any(), arg).Return(nil).AnyTimes()
+		check.Args(arg).Asserts(u, policy.ActionUpdatePersonal).Returns()
+	}))
+	s.Run("UpdateUserChatProviderKey", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
+		u := testutil.Fake(s.T(), faker, database.User{})
+		arg := database.UpdateUserChatProviderKeyParams{UserID: u.ID, ChatProviderID: uuid.New(), APIKey: "updated-api-key"}
+		key := testutil.Fake(s.T(), faker, database.UserChatProviderKey{UserID: u.ID, ChatProviderID: arg.ChatProviderID, APIKey: arg.APIKey})
+		dbm.EXPECT().GetUserByID(gomock.Any(), u.ID).Return(u, nil).AnyTimes()
+		dbm.EXPECT().UpdateUserChatProviderKey(gomock.Any(), arg).Return(key, nil).AnyTimes()
+		check.Args(arg).Asserts(u, policy.ActionUpdatePersonal).Returns(key)
+	}))
+	s.Run("UpsertUserChatProviderKey", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
+		u := testutil.Fake(s.T(), faker, database.User{})
+		arg := database.UpsertUserChatProviderKeyParams{UserID: u.ID, ChatProviderID: uuid.New(), APIKey: "upserted-api-key"}
+		key := testutil.Fake(s.T(), faker, database.UserChatProviderKey{UserID: u.ID, ChatProviderID: arg.ChatProviderID, APIKey: arg.APIKey})
+		dbm.EXPECT().GetUserByID(gomock.Any(), u.ID).Return(u, nil).AnyTimes()
+		dbm.EXPECT().UpsertUserChatProviderKey(gomock.Any(), arg).Return(key, nil).AnyTimes()
+		check.Args(arg).Asserts(u, policy.ActionUpdatePersonal).Returns(key)
+	}))
 	s.Run("UpdateUserChatCustomPrompt", s.Mocked(func(dbm *dbmock.MockStore, faker *gofakeit.Faker, check *expects) {
 		u := testutil.Fake(s.T(), faker, database.User{})
 		uc := database.UserConfig{UserID: u.ID, Key: "chat_custom_prompt", Value: "my custom prompt"}
@@ -3718,13 +3773,11 @@ func (s *MethodTestSuite) TestTailnetFunctions() {
 		check.Args(uuid.New()).
 			Asserts(rbac.ResourceTailnetCoordinator, policy.ActionRead)
 	}))
-	s.Run("GetTailnetTunnelPeerBindings", s.Subtest(func(_ database.Store, check *expects) {
-		check.Args(uuid.New()).
-			Asserts(rbac.ResourceTailnetCoordinator, policy.ActionRead)
+	s.Run("GetTailnetTunnelPeerBindingsBatch", s.Subtest(func(_ database.Store, check *expects) {
+		check.Args([]uuid.UUID{uuid.New()}).Asserts(rbac.ResourceTailnetCoordinator, policy.ActionRead)
 	}))
-	s.Run("GetTailnetTunnelPeerIDs", s.Subtest(func(_ database.Store, check *expects) {
-		check.Args(uuid.New()).
-			Asserts(rbac.ResourceTailnetCoordinator, policy.ActionRead)
+	s.Run("GetTailnetTunnelPeerIDsBatch", s.Subtest(func(_ database.Store, check *expects) {
+		check.Args([]uuid.UUID{uuid.New()}).Asserts(rbac.ResourceTailnetCoordinator, policy.ActionRead)
 	}))
 	s.Run("GetAllTailnetCoordinators", s.Subtest(func(_ database.Store, check *expects) {
 		check.Args().

coderd/database/dbgen/dbgen.go

@@ -76,7 +76,7 @@ func AuditLog(t testing.TB, db database.Store, seed database.AuditLog) database.
 }
 
 func ConnectionLog(t testing.TB, db database.Store, seed database.UpsertConnectionLogParams) database.ConnectionLog {
-	log, err := db.UpsertConnectionLog(genCtx, database.UpsertConnectionLogParams{
+	arg := database.UpsertConnectionLogParams{
 		ID:               takeFirst(seed.ID, uuid.New()),
 		Time:             takeFirst(seed.Time, dbtime.Now()),
 		OrganizationID:   takeFirst(seed.OrganizationID, uuid.New()),
@@ -89,7 +89,7 @@ func ConnectionLog(t testing.TB, db database.Store, seed database.UpsertConnecti
 			Int32: takeFirst(seed.Code.Int32, 0),
 			Valid: takeFirst(seed.Code.Valid, false),
 		},
-		Ip: pqtype.Inet{
+		IP: pqtype.Inet{
 			IPNet: net.IPNet{
 				IP:   net.IPv4(127, 0, 0, 1),
 				Mask: net.IPv4Mask(255, 255, 255, 255),
@@ -117,9 +117,53 @@ func ConnectionLog(t testing.TB, db database.Store, seed database.UpsertConnecti
 			Valid:  takeFirst(seed.DisconnectReason.Valid, false),
 		},
 		ConnectionStatus: takeFirst(seed.ConnectionStatus, database.ConnectionStatusConnected),
+	}
+
+	var disconnectTime sql.NullTime
+	if arg.ConnectionStatus == database.ConnectionStatusDisconnected {
+		disconnectTime = sql.NullTime{Time: arg.Time, Valid: true}
+	}
+
+	err := db.BatchUpsertConnectionLogs(genCtx, database.BatchUpsertConnectionLogsParams{
+		ID:               []uuid.UUID{arg.ID},
+		ConnectTime:      []time.Time{arg.Time},
+		OrganizationID:   []uuid.UUID{arg.OrganizationID},
+		WorkspaceOwnerID: []uuid.UUID{arg.WorkspaceOwnerID},
+		WorkspaceID:      []uuid.UUID{arg.WorkspaceID},
+		WorkspaceName:    []string{arg.WorkspaceName},
+		AgentName:        []string{arg.AgentName},
+		Type:             []database.ConnectionType{arg.Type},
+		Code:             []int32{arg.Code.Int32},
+		CodeValid:        []bool{arg.Code.Valid},
+		Ip:               []pqtype.Inet{arg.IP},
+		UserAgent:        []string{arg.UserAgent.String},
+		UserID:           []uuid.UUID{arg.UserID.UUID},
+		SlugOrPort:       []string{arg.SlugOrPort.String},
+		ConnectionID:     []uuid.UUID{arg.ConnectionID.UUID},
+		DisconnectReason: []string{arg.DisconnectReason.String},
+		DisconnectTime:   []time.Time{disconnectTime.Time},
 	})
 	require.NoError(t, err, "insert connection log")
-	return log
+
+	// Query back the actual row from the database. On upsert
+	// conflict the DB keeps the original row's ID, so we can't
+	// rely on arg.ID. Match on the conflict key for rows with a
+	// connection_id, or by primary key for NULL connection_id.
+	rows, err := db.GetConnectionLogsOffset(genCtx, database.GetConnectionLogsOffsetParams{})
+	require.NoError(t, err, "query connection logs")
+	for _, row := range rows {
+		if arg.ConnectionID.Valid {
+			if row.ConnectionLog.ConnectionID == arg.ConnectionID &&
+				row.ConnectionLog.WorkspaceID == arg.WorkspaceID &&
+				row.ConnectionLog.AgentName == arg.AgentName {
+				return row.ConnectionLog
+			}
+		} else if row.ConnectionLog.ID == arg.ID {
+			return row.ConnectionLog
+		}
+	}
+	require.Failf(t, "connection log not found", "id=%s", arg.ID)
+	return database.ConnectionLog{} // unreachable
 }
 
 func Template(t testing.TB, db database.Store, seed database.Template) database.Template {
@@ -1591,6 +1635,7 @@ func AIBridgeInterception(t testing.TB, db database.Store, seed database.InsertA
 		APIKeyID:                   seed.APIKeyID,
 		InitiatorID:                takeFirst(seed.InitiatorID, uuid.New()),
 		Provider:                   takeFirst(seed.Provider, "provider"),
+		ProviderName:               takeFirst(seed.ProviderName, "provider-name"),
 		Model:                      takeFirst(seed.Model, "model"),
 		Metadata:                   takeFirstSlice(seed.Metadata, json.RawMessage("{}")),
 		StartedAt:                  takeFirst(seed.StartedAt, dbtime.Now()),
@@ -1612,13 +1657,15 @@ func AIBridgeInterception(t testing.TB, db database.Store, seed database.InsertA
 
 func AIBridgeTokenUsage(t testing.TB, db database.Store, seed database.InsertAIBridgeTokenUsageParams) database.AIBridgeTokenUsage {
 	usage, err := db.InsertAIBridgeTokenUsage(genCtx, database.InsertAIBridgeTokenUsageParams{
-		ID:                 takeFirst(seed.ID, uuid.New()),
-		InterceptionID:     takeFirst(seed.InterceptionID, uuid.New()),
-		ProviderResponseID: takeFirst(seed.ProviderResponseID, "provider_response_id"),
-		InputTokens:        takeFirst(seed.InputTokens, 100),
-		OutputTokens:       takeFirst(seed.OutputTokens, 100),
-		Metadata:           takeFirstSlice(seed.Metadata, json.RawMessage("{}")),
-		CreatedAt:          takeFirst(seed.CreatedAt, dbtime.Now()),
+		ID:                    takeFirst(seed.ID, uuid.New()),
+		InterceptionID:        takeFirst(seed.InterceptionID, uuid.New()),
+		ProviderResponseID:    takeFirst(seed.ProviderResponseID, "provider_response_id"),
+		InputTokens:           takeFirst(seed.InputTokens, 100),
+		OutputTokens:          takeFirst(seed.OutputTokens, 100),
+		CacheReadInputTokens:  seed.CacheReadInputTokens,
+		CacheWriteInputTokens: seed.CacheWriteInputTokens,
+		Metadata:              takeFirstSlice(seed.Metadata, json.RawMessage("{}")),
+		CreatedAt:             takeFirst(seed.CreatedAt, dbtime.Now()),
 	})
 	require.NoError(t, err, "insert aibridge token usage")
 	return usage

coderd/database/dbmetrics/querymetrics.go

@@ -160,12 +160,12 @@ func (m queryMetricsStore) AllUserIDs(ctx context.Context, includeSystem bool) (
 	return r0, r1
 }
 
-func (m queryMetricsStore) ArchiveChatByID(ctx context.Context, id uuid.UUID) error {
+func (m queryMetricsStore) ArchiveChatByID(ctx context.Context, id uuid.UUID) ([]database.Chat, error) {
 	start := time.Now()
-	r0 := m.s.ArchiveChatByID(ctx, id)
+	r0, r1 := m.s.ArchiveChatByID(ctx, id)
 	m.queryLatencies.WithLabelValues("ArchiveChatByID").Observe(time.Since(start).Seconds())
 	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "ArchiveChatByID").Inc()
-	return r0
+	return r0, r1
 }
 
 func (m queryMetricsStore) ArchiveUnusedTemplateVersions(ctx context.Context, arg database.ArchiveUnusedTemplateVersionsParams) ([]uuid.UUID, error) {
@@ -208,6 +208,14 @@ func (m queryMetricsStore) BatchUpdateWorkspaceNextStartAt(ctx context.Context,
 	return r0
 }
 
+func (m queryMetricsStore) BatchUpsertConnectionLogs(ctx context.Context, arg database.BatchUpsertConnectionLogsParams) error {
+	start := time.Now()
+	r0 := m.s.BatchUpsertConnectionLogs(ctx, arg)
+	m.queryLatencies.WithLabelValues("BatchUpsertConnectionLogs").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "BatchUpsertConnectionLogs").Inc()
+	return r0
+}
+
 func (m queryMetricsStore) BulkMarkNotificationMessagesFailed(ctx context.Context, arg database.BulkMarkNotificationMessagesFailedParams) (int64, error) {
 	start := time.Now()
 	r0, r1 := m.s.BulkMarkNotificationMessagesFailed(ctx, arg)
@@ -696,6 +704,14 @@ func (m queryMetricsStore) DeleteUserChatCompactionThreshold(ctx context.Context
 	return r0
 }
 
+func (m queryMetricsStore) DeleteUserChatProviderKey(ctx context.Context, arg database.DeleteUserChatProviderKeyParams) error {
+	start := time.Now()
+	r0 := m.s.DeleteUserChatProviderKey(ctx, arg)
+	m.queryLatencies.WithLabelValues("DeleteUserChatProviderKey").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "DeleteUserChatProviderKey").Inc()
+	return r0
+}
+
 func (m queryMetricsStore) DeleteUserSecret(ctx context.Context, id uuid.UUID) error {
 	start := time.Now()
 	r0 := m.s.DeleteUserSecret(ctx, id)
@@ -1112,6 +1128,14 @@ func (m queryMetricsStore) GetChatFileByID(ctx context.Context, id uuid.UUID) (d
 	return r0, r1
 }
 
+func (m queryMetricsStore) GetChatFileMetadataByChatID(ctx context.Context, chatID uuid.UUID) ([]database.GetChatFileMetadataByChatIDRow, error) {
+	start := time.Now()
+	r0, r1 := m.s.GetChatFileMetadataByChatID(ctx, chatID)
+	m.queryLatencies.WithLabelValues("GetChatFileMetadataByChatID").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "GetChatFileMetadataByChatID").Inc()
+	return r0, r1
+}
+
 func (m queryMetricsStore) GetChatFilesByIDs(ctx context.Context, ids []uuid.UUID) ([]database.ChatFile, error) {
 	start := time.Now()
 	r0, r1 := m.s.GetChatFilesByIDs(ctx, ids)
@@ -2208,19 +2232,19 @@ func (m queryMetricsStore) GetTailnetPeers(ctx context.Context, id uuid.UUID) ([
 	return r0, r1
 }
 
-func (m queryMetricsStore) GetTailnetTunnelPeerBindings(ctx context.Context, srcID uuid.UUID) ([]database.GetTailnetTunnelPeerBindingsRow, error) {
+func (m queryMetricsStore) GetTailnetTunnelPeerBindingsBatch(ctx context.Context, ids []uuid.UUID) ([]database.GetTailnetTunnelPeerBindingsBatchRow, error) {
 	start := time.Now()
-	r0, r1 := m.s.GetTailnetTunnelPeerBindings(ctx, srcID)
-	m.queryLatencies.WithLabelValues("GetTailnetTunnelPeerBindings").Observe(time.Since(start).Seconds())
-	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "GetTailnetTunnelPeerBindings").Inc()
+	r0, r1 := m.s.GetTailnetTunnelPeerBindingsBatch(ctx, ids)
+	m.queryLatencies.WithLabelValues("GetTailnetTunnelPeerBindingsBatch").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "GetTailnetTunnelPeerBindingsBatch").Inc()
 	return r0, r1
 }
 
-func (m queryMetricsStore) GetTailnetTunnelPeerIDs(ctx context.Context, srcID uuid.UUID) ([]database.GetTailnetTunnelPeerIDsRow, error) {
+func (m queryMetricsStore) GetTailnetTunnelPeerIDsBatch(ctx context.Context, ids []uuid.UUID) ([]database.GetTailnetTunnelPeerIDsBatchRow, error) {
 	start := time.Now()
-	r0, r1 := m.s.GetTailnetTunnelPeerIDs(ctx, srcID)
-	m.queryLatencies.WithLabelValues("GetTailnetTunnelPeerIDs").Observe(time.Since(start).Seconds())
-	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "GetTailnetTunnelPeerIDs").Inc()
+	r0, r1 := m.s.GetTailnetTunnelPeerIDsBatch(ctx, ids)
+	m.queryLatencies.WithLabelValues("GetTailnetTunnelPeerIDsBatch").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "GetTailnetTunnelPeerIDsBatch").Inc()
 	return r0, r1
 }
 
@@ -2528,6 +2552,14 @@ func (m queryMetricsStore) GetUserChatCustomPrompt(ctx context.Context, userID u
 	return r0, r1
 }
 
+func (m queryMetricsStore) GetUserChatProviderKeys(ctx context.Context, userID uuid.UUID) ([]database.UserChatProviderKey, error) {
+	start := time.Now()
+	r0, r1 := m.s.GetUserChatProviderKeys(ctx, userID)
+	m.queryLatencies.WithLabelValues("GetUserChatProviderKeys").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "GetUserChatProviderKeys").Inc()
+	return r0, r1
+}
+
 func (m queryMetricsStore) GetUserChatSpendInPeriod(ctx context.Context, arg database.GetUserChatSpendInPeriodParams) (int64, error) {
 	start := time.Now()
 	r0, r1 := m.s.GetUserChatSpendInPeriod(ctx, arg)
@@ -3752,6 +3784,14 @@ func (m queryMetricsStore) InsertWorkspaceResourceMetadata(ctx context.Context,
 	return r0, r1
 }
 
+func (m queryMetricsStore) LinkChatFiles(ctx context.Context, arg database.LinkChatFilesParams) (int32, error) {
+	start := time.Now()
+	r0, r1 := m.s.LinkChatFiles(ctx, arg)
+	m.queryLatencies.WithLabelValues("LinkChatFiles").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "LinkChatFiles").Inc()
+	return r0, r1
+}
+
 func (m queryMetricsStore) ListAIBridgeClients(ctx context.Context, arg database.ListAIBridgeClientsParams) ([]string, error) {
 	start := time.Now()
 	r0, r1 := m.s.ListAIBridgeClients(ctx, arg)
@@ -4024,12 +4064,12 @@ func (m queryMetricsStore) TryAcquireLock(ctx context.Context, pgTryAdvisoryXact
 	return r0, r1
 }
 
-func (m queryMetricsStore) UnarchiveChatByID(ctx context.Context, id uuid.UUID) error {
+func (m queryMetricsStore) UnarchiveChatByID(ctx context.Context, id uuid.UUID) ([]database.Chat, error) {
 	start := time.Now()
-	r0 := m.s.UnarchiveChatByID(ctx, id)
+	r0, r1 := m.s.UnarchiveChatByID(ctx, id)
 	m.queryLatencies.WithLabelValues("UnarchiveChatByID").Observe(time.Since(start).Seconds())
 	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "UnarchiveChatByID").Inc()
-	return r0
+	return r0, r1
 }
 
 func (m queryMetricsStore) UnarchiveTemplateVersion(ctx context.Context, arg database.UnarchiveTemplateVersionParams) error {
@@ -4096,11 +4136,11 @@ func (m queryMetricsStore) UpdateChatByID(ctx context.Context, arg database.Upda
 	return r0, r1
 }
 
-func (m queryMetricsStore) UpdateChatHeartbeat(ctx context.Context, arg database.UpdateChatHeartbeatParams) (int64, error) {
+func (m queryMetricsStore) UpdateChatHeartbeats(ctx context.Context, arg database.UpdateChatHeartbeatsParams) ([]uuid.UUID, error) {
 	start := time.Now()
-	r0, r1 := m.s.UpdateChatHeartbeat(ctx, arg)
-	m.queryLatencies.WithLabelValues("UpdateChatHeartbeat").Observe(time.Since(start).Seconds())
-	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "UpdateChatHeartbeat").Inc()
+	r0, r1 := m.s.UpdateChatHeartbeats(ctx, arg)
+	m.queryLatencies.WithLabelValues("UpdateChatHeartbeats").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "UpdateChatHeartbeats").Inc()
 	return r0, r1
 }
 
@@ -4560,6 +4600,14 @@ func (m queryMetricsStore) UpdateUserChatCustomPrompt(ctx context.Context, arg d
 	return r0, r1
 }
 
+func (m queryMetricsStore) UpdateUserChatProviderKey(ctx context.Context, arg database.UpdateUserChatProviderKeyParams) (database.UserChatProviderKey, error) {
+	start := time.Now()
+	r0, r1 := m.s.UpdateUserChatProviderKey(ctx, arg)
+	m.queryLatencies.WithLabelValues("UpdateUserChatProviderKey").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "UpdateUserChatProviderKey").Inc()
+	return r0, r1
+}
+
 func (m queryMetricsStore) UpdateUserDeletedByID(ctx context.Context, id uuid.UUID) error {
 	start := time.Now()
 	r0 := m.s.UpdateUserDeletedByID(ctx, id)
@@ -5000,14 +5048,6 @@ func (m queryMetricsStore) UpsertChatWorkspaceTTL(ctx context.Context, workspace
 	return r0
 }
 
-func (m queryMetricsStore) UpsertConnectionLog(ctx context.Context, arg database.UpsertConnectionLogParams) (database.ConnectionLog, error) {
-	start := time.Now()
-	r0, r1 := m.s.UpsertConnectionLog(ctx, arg)
-	m.queryLatencies.WithLabelValues("UpsertConnectionLog").Observe(time.Since(start).Seconds())
-	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "UpsertConnectionLog").Inc()
-	return r0, r1
-}
-
 func (m queryMetricsStore) UpsertDefaultProxy(ctx context.Context, arg database.UpsertDefaultProxyParams) error {
 	start := time.Now()
 	r0 := m.s.UpsertDefaultProxy(ctx, arg)
@@ -5152,6 +5192,14 @@ func (m queryMetricsStore) UpsertTemplateUsageStats(ctx context.Context) error {
 	return r0
 }
 
+func (m queryMetricsStore) UpsertUserChatProviderKey(ctx context.Context, arg database.UpsertUserChatProviderKeyParams) (database.UserChatProviderKey, error) {
+	start := time.Now()
+	r0, r1 := m.s.UpsertUserChatProviderKey(ctx, arg)
+	m.queryLatencies.WithLabelValues("UpsertUserChatProviderKey").Observe(time.Since(start).Seconds())
+	m.queryCounts.WithLabelValues(httpmw.ExtractHTTPRoute(ctx), httpmw.ExtractHTTPMethod(ctx), "UpsertUserChatProviderKey").Inc()
+	return r0, r1
+}
+
 func (m queryMetricsStore) UpsertWebpushVAPIDKeys(ctx context.Context, arg database.UpsertWebpushVAPIDKeysParams) error {
 	start := time.Now()
 	r0 := m.s.UpsertWebpushVAPIDKeys(ctx, arg)

coderd/database/dbmock/dbmock.go

@@ -148,11 +148,12 @@ func (mr *MockStoreMockRecorder) AllUserIDs(ctx, includeSystem any) *gomock.Call
 }
 
 // ArchiveChatByID mocks base method.
-func (m *MockStore) ArchiveChatByID(ctx context.Context, id uuid.UUID) error {
+func (m *MockStore) ArchiveChatByID(ctx context.Context, id uuid.UUID) ([]database.Chat, error) {
 	m.ctrl.T.Helper()
 	ret := m.ctrl.Call(m, "ArchiveChatByID", ctx, id)
-	ret0, _ := ret[0].(error)
-	return ret0
+	ret0, _ := ret[0].([]database.Chat)
+	ret1, _ := ret[1].(error)
+	return ret0, ret1
 }
 
 // ArchiveChatByID indicates an expected call of ArchiveChatByID.
@@ -232,6 +233,20 @@ func (mr *MockStoreMockRecorder) BatchUpdateWorkspaceNextStartAt(ctx, arg any) *
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "BatchUpdateWorkspaceNextStartAt", reflect.TypeOf((*MockStore)(nil).BatchUpdateWorkspaceNextStartAt), ctx, arg)
 }
 
+// BatchUpsertConnectionLogs mocks base method.
+func (m *MockStore) BatchUpsertConnectionLogs(ctx context.Context, arg database.BatchUpsertConnectionLogsParams) error {
+	m.ctrl.T.Helper()
+	ret := m.ctrl.Call(m, "BatchUpsertConnectionLogs", ctx, arg)
+	ret0, _ := ret[0].(error)
+	return ret0
+}
+
+// BatchUpsertConnectionLogs indicates an expected call of BatchUpsertConnectionLogs.
+func (mr *MockStoreMockRecorder) BatchUpsertConnectionLogs(ctx, arg any) *gomock.Call {
+	mr.mock.ctrl.T.Helper()
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "BatchUpsertConnectionLogs", reflect.TypeOf((*MockStore)(nil).BatchUpsertConnectionLogs), ctx, arg)
+}
+
 // BulkMarkNotificationMessagesFailed mocks base method.
 func (m *MockStore) BulkMarkNotificationMessagesFailed(ctx context.Context, arg database.BulkMarkNotificationMessagesFailedParams) (int64, error) {
 	m.ctrl.T.Helper()
@@ -1170,6 +1185,20 @@ func (mr *MockStoreMockRecorder) DeleteUserChatCompactionThreshold(ctx, arg any)
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "DeleteUserChatCompactionThreshold", reflect.TypeOf((*MockStore)(nil).DeleteUserChatCompactionThreshold), ctx, arg)
 }
 
+// DeleteUserChatProviderKey mocks base method.
+func (m *MockStore) DeleteUserChatProviderKey(ctx context.Context, arg database.DeleteUserChatProviderKeyParams) error {
+	m.ctrl.T.Helper()
+	ret := m.ctrl.Call(m, "DeleteUserChatProviderKey", ctx, arg)
+	ret0, _ := ret[0].(error)
+	return ret0
+}
+
+// DeleteUserChatProviderKey indicates an expected call of DeleteUserChatProviderKey.
+func (mr *MockStoreMockRecorder) DeleteUserChatProviderKey(ctx, arg any) *gomock.Call {
+	mr.mock.ctrl.T.Helper()
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "DeleteUserChatProviderKey", reflect.TypeOf((*MockStore)(nil).DeleteUserChatProviderKey), ctx, arg)
+}
+
 // DeleteUserSecret mocks base method.
 func (m *MockStore) DeleteUserSecret(ctx context.Context, id uuid.UUID) error {
 	m.ctrl.T.Helper()
@@ -2043,6 +2072,21 @@ func (mr *MockStoreMockRecorder) GetChatFileByID(ctx, id any) *gomock.Call {
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetChatFileByID", reflect.TypeOf((*MockStore)(nil).GetChatFileByID), ctx, id)
 }
 
+// GetChatFileMetadataByChatID mocks base method.
+func (m *MockStore) GetChatFileMetadataByChatID(ctx context.Context, chatID uuid.UUID) ([]database.GetChatFileMetadataByChatIDRow, error) {
+	m.ctrl.T.Helper()
+	ret := m.ctrl.Call(m, "GetChatFileMetadataByChatID", ctx, chatID)
+	ret0, _ := ret[0].([]database.GetChatFileMetadataByChatIDRow)
+	ret1, _ := ret[1].(error)
+	return ret0, ret1
+}
+
+// GetChatFileMetadataByChatID indicates an expected call of GetChatFileMetadataByChatID.
+func (mr *MockStoreMockRecorder) GetChatFileMetadataByChatID(ctx, chatID any) *gomock.Call {
+	mr.mock.ctrl.T.Helper()
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetChatFileMetadataByChatID", reflect.TypeOf((*MockStore)(nil).GetChatFileMetadataByChatID), ctx, chatID)
+}
+
 // GetChatFilesByIDs mocks base method.
 func (m *MockStore) GetChatFilesByIDs(ctx context.Context, ids []uuid.UUID) ([]database.ChatFile, error) {
 	m.ctrl.T.Helper()
@@ -4098,34 +4142,34 @@ func (mr *MockStoreMockRecorder) GetTailnetPeers(ctx, id any) *gomock.Call {
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetTailnetPeers", reflect.TypeOf((*MockStore)(nil).GetTailnetPeers), ctx, id)
 }
 
-// GetTailnetTunnelPeerBindings mocks base method.
-func (m *MockStore) GetTailnetTunnelPeerBindings(ctx context.Context, srcID uuid.UUID) ([]database.GetTailnetTunnelPeerBindingsRow, error) {
+// GetTailnetTunnelPeerBindingsBatch mocks base method.
+func (m *MockStore) GetTailnetTunnelPeerBindingsBatch(ctx context.Context, ids []uuid.UUID) ([]database.GetTailnetTunnelPeerBindingsBatchRow, error) {
 	m.ctrl.T.Helper()
-	ret := m.ctrl.Call(m, "GetTailnetTunnelPeerBindings", ctx, srcID)
-	ret0, _ := ret[0].([]database.GetTailnetTunnelPeerBindingsRow)
+	ret := m.ctrl.Call(m, "GetTailnetTunnelPeerBindingsBatch", ctx, ids)
+	ret0, _ := ret[0].([]database.GetTailnetTunnelPeerBindingsBatchRow)
 	ret1, _ := ret[1].(error)
 	return ret0, ret1
 }
 
-// GetTailnetTunnelPeerBindings indicates an expected call of GetTailnetTunnelPeerBindings.
-func (mr *MockStoreMockRecorder) GetTailnetTunnelPeerBindings(ctx, srcID any) *gomock.Call {
+// GetTailnetTunnelPeerBindingsBatch indicates an expected call of GetTailnetTunnelPeerBindingsBatch.
+func (mr *MockStoreMockRecorder) GetTailnetTunnelPeerBindingsBatch(ctx, ids any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
-	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetTailnetTunnelPeerBindings", reflect.TypeOf((*MockStore)(nil).GetTailnetTunnelPeerBindings), ctx, srcID)
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetTailnetTunnelPeerBindingsBatch", reflect.TypeOf((*MockStore)(nil).GetTailnetTunnelPeerBindingsBatch), ctx, ids)
 }
 
-// GetTailnetTunnelPeerIDs mocks base method.
-func (m *MockStore) GetTailnetTunnelPeerIDs(ctx context.Context, srcID uuid.UUID) ([]database.GetTailnetTunnelPeerIDsRow, error) {
+// GetTailnetTunnelPeerIDsBatch mocks base method.
+func (m *MockStore) GetTailnetTunnelPeerIDsBatch(ctx context.Context, ids []uuid.UUID) ([]database.GetTailnetTunnelPeerIDsBatchRow, error) {
 	m.ctrl.T.Helper()
-	ret := m.ctrl.Call(m, "GetTailnetTunnelPeerIDs", ctx, srcID)
-	ret0, _ := ret[0].([]database.GetTailnetTunnelPeerIDsRow)
+	ret := m.ctrl.Call(m, "GetTailnetTunnelPeerIDsBatch", ctx, ids)
+	ret0, _ := ret[0].([]database.GetTailnetTunnelPeerIDsBatchRow)
 	ret1, _ := ret[1].(error)
 	return ret0, ret1
 }
 
-// GetTailnetTunnelPeerIDs indicates an expected call of GetTailnetTunnelPeerIDs.
-func (mr *MockStoreMockRecorder) GetTailnetTunnelPeerIDs(ctx, srcID any) *gomock.Call {
+// GetTailnetTunnelPeerIDsBatch indicates an expected call of GetTailnetTunnelPeerIDsBatch.
+func (mr *MockStoreMockRecorder) GetTailnetTunnelPeerIDsBatch(ctx, ids any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
-	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetTailnetTunnelPeerIDs", reflect.TypeOf((*MockStore)(nil).GetTailnetTunnelPeerIDs), ctx, srcID)
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetTailnetTunnelPeerIDsBatch", reflect.TypeOf((*MockStore)(nil).GetTailnetTunnelPeerIDsBatch), ctx, ids)
 }
 
 // GetTaskByID mocks base method.
@@ -4728,6 +4772,21 @@ func (mr *MockStoreMockRecorder) GetUserChatCustomPrompt(ctx, userID any) *gomoc
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetUserChatCustomPrompt", reflect.TypeOf((*MockStore)(nil).GetUserChatCustomPrompt), ctx, userID)
 }
 
+// GetUserChatProviderKeys mocks base method.
+func (m *MockStore) GetUserChatProviderKeys(ctx context.Context, userID uuid.UUID) ([]database.UserChatProviderKey, error) {
+	m.ctrl.T.Helper()
+	ret := m.ctrl.Call(m, "GetUserChatProviderKeys", ctx, userID)
+	ret0, _ := ret[0].([]database.UserChatProviderKey)
+	ret1, _ := ret[1].(error)
+	return ret0, ret1
+}
+
+// GetUserChatProviderKeys indicates an expected call of GetUserChatProviderKeys.
+func (mr *MockStoreMockRecorder) GetUserChatProviderKeys(ctx, userID any) *gomock.Call {
+	mr.mock.ctrl.T.Helper()
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetUserChatProviderKeys", reflect.TypeOf((*MockStore)(nil).GetUserChatProviderKeys), ctx, userID)
+}
+
 // GetUserChatSpendInPeriod mocks base method.
 func (m *MockStore) GetUserChatSpendInPeriod(ctx context.Context, arg database.GetUserChatSpendInPeriodParams) (int64, error) {
 	m.ctrl.T.Helper()
@@ -7022,6 +7081,21 @@ func (mr *MockStoreMockRecorder) InsertWorkspaceResourceMetadata(ctx, arg any) *
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "InsertWorkspaceResourceMetadata", reflect.TypeOf((*MockStore)(nil).InsertWorkspaceResourceMetadata), ctx, arg)
 }
 
+// LinkChatFiles mocks base method.
+func (m *MockStore) LinkChatFiles(ctx context.Context, arg database.LinkChatFilesParams) (int32, error) {
+	m.ctrl.T.Helper()
+	ret := m.ctrl.Call(m, "LinkChatFiles", ctx, arg)
+	ret0, _ := ret[0].(int32)
+	ret1, _ := ret[1].(error)
+	return ret0, ret1
+}
+
+// LinkChatFiles indicates an expected call of LinkChatFiles.
+func (mr *MockStoreMockRecorder) LinkChatFiles(ctx, arg any) *gomock.Call {
+	mr.mock.ctrl.T.Helper()
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "LinkChatFiles", reflect.TypeOf((*MockStore)(nil).LinkChatFiles), ctx, arg)
+}
+
 // ListAIBridgeClients mocks base method.
 func (m *MockStore) ListAIBridgeClients(ctx context.Context, arg database.ListAIBridgeClientsParams) ([]string, error) {
 	m.ctrl.T.Helper()
@@ -7632,11 +7706,12 @@ func (mr *MockStoreMockRecorder) TryAcquireLock(ctx, pgTryAdvisoryXactLock any)
 }
 
 // UnarchiveChatByID mocks base method.
-func (m *MockStore) UnarchiveChatByID(ctx context.Context, id uuid.UUID) error {
+func (m *MockStore) UnarchiveChatByID(ctx context.Context, id uuid.UUID) ([]database.Chat, error) {
 	m.ctrl.T.Helper()
 	ret := m.ctrl.Call(m, "UnarchiveChatByID", ctx, id)
-	ret0, _ := ret[0].(error)
-	return ret0
+	ret0, _ := ret[0].([]database.Chat)
+	ret1, _ := ret[1].(error)
+	return ret0, ret1
 }
 
 // UnarchiveChatByID indicates an expected call of UnarchiveChatByID.
@@ -7760,19 +7835,19 @@ func (mr *MockStoreMockRecorder) UpdateChatByID(ctx, arg any) *gomock.Call {
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpdateChatByID", reflect.TypeOf((*MockStore)(nil).UpdateChatByID), ctx, arg)
 }
 
-// UpdateChatHeartbeat mocks base method.
-func (m *MockStore) UpdateChatHeartbeat(ctx context.Context, arg database.UpdateChatHeartbeatParams) (int64, error) {
+// UpdateChatHeartbeats mocks base method.
+func (m *MockStore) UpdateChatHeartbeats(ctx context.Context, arg database.UpdateChatHeartbeatsParams) ([]uuid.UUID, error) {
 	m.ctrl.T.Helper()
-	ret := m.ctrl.Call(m, "UpdateChatHeartbeat", ctx, arg)
-	ret0, _ := ret[0].(int64)
+	ret := m.ctrl.Call(m, "UpdateChatHeartbeats", ctx, arg)
+	ret0, _ := ret[0].([]uuid.UUID)
 	ret1, _ := ret[1].(error)
 	return ret0, ret1
 }
 
-// UpdateChatHeartbeat indicates an expected call of UpdateChatHeartbeat.
-func (mr *MockStoreMockRecorder) UpdateChatHeartbeat(ctx, arg any) *gomock.Call {
+// UpdateChatHeartbeats indicates an expected call of UpdateChatHeartbeats.
+func (mr *MockStoreMockRecorder) UpdateChatHeartbeats(ctx, arg any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
-	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpdateChatHeartbeat", reflect.TypeOf((*MockStore)(nil).UpdateChatHeartbeat), ctx, arg)
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpdateChatHeartbeats", reflect.TypeOf((*MockStore)(nil).UpdateChatHeartbeats), ctx, arg)
 }
 
 // UpdateChatLabelsByID mocks base method.
@@ -8603,6 +8678,21 @@ func (mr *MockStoreMockRecorder) UpdateUserChatCustomPrompt(ctx, arg any) *gomoc
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpdateUserChatCustomPrompt", reflect.TypeOf((*MockStore)(nil).UpdateUserChatCustomPrompt), ctx, arg)
 }
 
+// UpdateUserChatProviderKey mocks base method.
+func (m *MockStore) UpdateUserChatProviderKey(ctx context.Context, arg database.UpdateUserChatProviderKeyParams) (database.UserChatProviderKey, error) {
+	m.ctrl.T.Helper()
+	ret := m.ctrl.Call(m, "UpdateUserChatProviderKey", ctx, arg)
+	ret0, _ := ret[0].(database.UserChatProviderKey)
+	ret1, _ := ret[1].(error)
+	return ret0, ret1
+}
+
+// UpdateUserChatProviderKey indicates an expected call of UpdateUserChatProviderKey.
+func (mr *MockStoreMockRecorder) UpdateUserChatProviderKey(ctx, arg any) *gomock.Call {
+	mr.mock.ctrl.T.Helper()
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpdateUserChatProviderKey", reflect.TypeOf((*MockStore)(nil).UpdateUserChatProviderKey), ctx, arg)
+}
+
 // UpdateUserDeletedByID mocks base method.
 func (m *MockStore) UpdateUserDeletedByID(ctx context.Context, id uuid.UUID) error {
 	m.ctrl.T.Helper()
@@ -9396,21 +9486,6 @@ func (mr *MockStoreMockRecorder) UpsertChatWorkspaceTTL(ctx, workspaceTtl any) *
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpsertChatWorkspaceTTL", reflect.TypeOf((*MockStore)(nil).UpsertChatWorkspaceTTL), ctx, workspaceTtl)
 }
 
-// UpsertConnectionLog mocks base method.
-func (m *MockStore) UpsertConnectionLog(ctx context.Context, arg database.UpsertConnectionLogParams) (database.ConnectionLog, error) {
-	m.ctrl.T.Helper()
-	ret := m.ctrl.Call(m, "UpsertConnectionLog", ctx, arg)
-	ret0, _ := ret[0].(database.ConnectionLog)
-	ret1, _ := ret[1].(error)
-	return ret0, ret1
-}
-
-// UpsertConnectionLog indicates an expected call of UpsertConnectionLog.
-func (mr *MockStoreMockRecorder) UpsertConnectionLog(ctx, arg any) *gomock.Call {
-	mr.mock.ctrl.T.Helper()
-	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpsertConnectionLog", reflect.TypeOf((*MockStore)(nil).UpsertConnectionLog), ctx, arg)
-}
-
 // UpsertDefaultProxy mocks base method.
 func (m *MockStore) UpsertDefaultProxy(ctx context.Context, arg database.UpsertDefaultProxyParams) error {
 	m.ctrl.T.Helper()
@@ -9669,6 +9744,21 @@ func (mr *MockStoreMockRecorder) UpsertTemplateUsageStats(ctx any) *gomock.Call
 	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpsertTemplateUsageStats", reflect.TypeOf((*MockStore)(nil).UpsertTemplateUsageStats), ctx)
 }
 
+// UpsertUserChatProviderKey mocks base method.
+func (m *MockStore) UpsertUserChatProviderKey(ctx context.Context, arg database.UpsertUserChatProviderKeyParams) (database.UserChatProviderKey, error) {
+	m.ctrl.T.Helper()
+	ret := m.ctrl.Call(m, "UpsertUserChatProviderKey", ctx, arg)
+	ret0, _ := ret[0].(database.UserChatProviderKey)
+	ret1, _ := ret[1].(error)
+	return ret0, ret1
+}
+
+// UpsertUserChatProviderKey indicates an expected call of UpsertUserChatProviderKey.
+func (mr *MockStoreMockRecorder) UpsertUserChatProviderKey(ctx, arg any) *gomock.Call {
+	mr.mock.ctrl.T.Helper()
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "UpsertUserChatProviderKey", reflect.TypeOf((*MockStore)(nil).UpsertUserChatProviderKey), ctx, arg)
+}
+
 // UpsertWebpushVAPIDKeys mocks base method.
 func (m *MockStore) UpsertWebpushVAPIDKeys(ctx context.Context, arg database.UpsertWebpushVAPIDKeysParams) error {
 	m.ctrl.T.Helper()

coderd/database/dump.sql

@@ -1100,7 +1100,8 @@ CREATE TABLE aibridge_interceptions (
     thread_parent_id uuid,
     thread_root_id uuid,
     client_session_id character varying(256),
-    session_id text GENERATED ALWAYS AS (COALESCE(client_session_id, ((thread_root_id)::text)::character varying, ((id)::text)::character varying)) STORED NOT NULL
+    session_id text GENERATED ALWAYS AS (COALESCE(client_session_id, ((thread_root_id)::text)::character varying, ((id)::text)::character varying)) STORED NOT NULL,
+    provider_name text DEFAULT ''::text NOT NULL
 );
 
 COMMENT ON TABLE aibridge_interceptions IS 'Audit log of requests intercepted by AI Bridge';
@@ -1115,6 +1116,8 @@ COMMENT ON COLUMN aibridge_interceptions.client_session_id IS 'The session ID su
 
 COMMENT ON COLUMN aibridge_interceptions.session_id IS 'Groups related interceptions into a logical session. Determined by a priority chain: (1) client_session_id — an explicit session identifier supplied by the calling client (e.g. Claude Code); (2) thread_root_id — the root of an agentic thread detected by Bridge through tool-call correlation, used when the client does not supply its own session ID; (3) id — the interception''s own ID, used as a last resort so every interception belongs to exactly one session even if it is standalone. This is a generated column stored on disk so it can be indexed and joined without recomputing the COALESCE on every query.';
 
+COMMENT ON COLUMN aibridge_interceptions.provider_name IS 'The provider instance name which may differ from provider when multiple instances of the same provider type exist.';
+
 CREATE TABLE aibridge_model_thoughts (
     interception_id uuid NOT NULL,
     content text NOT NULL,
@@ -1131,7 +1134,9 @@ CREATE TABLE aibridge_token_usages (
     input_tokens bigint NOT NULL,
     output_tokens bigint NOT NULL,
     metadata jsonb,
-    created_at timestamp with time zone NOT NULL
+    created_at timestamp with time zone NOT NULL,
+    cache_read_input_tokens bigint DEFAULT 0 NOT NULL,
+    cache_write_input_tokens bigint DEFAULT 0 NOT NULL
 );
 
 COMMENT ON TABLE aibridge_token_usages IS 'Audit log of tokens used by intercepted requests in AI Bridge';
@@ -1264,6 +1269,11 @@ CREATE TABLE chat_diff_statuses (
     head_branch text
 );
 
+CREATE TABLE chat_file_links (
+    chat_id uuid NOT NULL,
+    file_id uuid NOT NULL
+);
+
 CREATE TABLE chat_files (
     id uuid DEFAULT gen_random_uuid() NOT NULL,
     owner_id uuid NOT NULL,
@@ -1338,7 +1348,11 @@ CREATE TABLE chat_providers (
     created_at timestamp with time zone DEFAULT now() NOT NULL,
     updated_at timestamp with time zone DEFAULT now() NOT NULL,
     base_url text DEFAULT ''::text NOT NULL,
-    CONSTRAINT chat_providers_provider_check CHECK ((provider = ANY (ARRAY['anthropic'::text, 'azure'::text, 'bedrock'::text, 'google'::text, 'openai'::text, 'openai-compat'::text, 'openrouter'::text, 'vercel'::text])))
+    central_api_key_enabled boolean DEFAULT true NOT NULL,
+    allow_user_api_key boolean DEFAULT false NOT NULL,
+    allow_central_api_key_fallback boolean DEFAULT false NOT NULL,
+    CONSTRAINT chat_providers_provider_check CHECK ((provider = ANY (ARRAY['anthropic'::text, 'azure'::text, 'bedrock'::text, 'google'::text, 'openai'::text, 'openai-compat'::text, 'openrouter'::text, 'vercel'::text]))),
+    CONSTRAINT valid_credential_policy CHECK (((central_api_key_enabled OR allow_user_api_key) AND ((NOT allow_central_api_key_fallback) OR (central_api_key_enabled AND allow_user_api_key))))
 );
 
 COMMENT ON COLUMN chat_providers.api_key_key_id IS 'The ID of the key used to encrypt the provider API key. If this is NULL, the API key is not encrypted';
@@ -2749,6 +2763,17 @@ COMMENT ON TABLE usage_events_daily IS 'usage_events_daily is a daily rollup of
 
 COMMENT ON COLUMN usage_events_daily.day IS 'The date of the summed usage events, always in UTC.';
 
+CREATE TABLE user_chat_provider_keys (
+    id uuid DEFAULT gen_random_uuid() NOT NULL,
+    user_id uuid NOT NULL,
+    chat_provider_id uuid NOT NULL,
+    api_key text NOT NULL,
+    api_key_key_id text,
+    created_at timestamp with time zone DEFAULT now() NOT NULL,
+    updated_at timestamp with time zone DEFAULT now() NOT NULL,
+    CONSTRAINT user_chat_provider_keys_api_key_check CHECK ((api_key <> ''::text))
+);
+
 CREATE TABLE user_configs (
     user_id uuid NOT NULL,
     key character varying(256) NOT NULL,
@@ -2790,7 +2815,8 @@ CREATE TABLE user_secrets (
     env_name text DEFAULT ''::text NOT NULL,
     file_path text DEFAULT ''::text NOT NULL,
     created_at timestamp with time zone DEFAULT CURRENT_TIMESTAMP NOT NULL,
-    updated_at timestamp with time zone DEFAULT CURRENT_TIMESTAMP NOT NULL
+    updated_at timestamp with time zone DEFAULT CURRENT_TIMESTAMP NOT NULL,
+    value_key_id text
 );
 
 CREATE TABLE user_status_changes (
@@ -3323,6 +3349,9 @@ ALTER TABLE ONLY boundary_usage_stats
 ALTER TABLE ONLY chat_diff_statuses
     ADD CONSTRAINT chat_diff_statuses_pkey PRIMARY KEY (chat_id);
 
+ALTER TABLE ONLY chat_file_links
+    ADD CONSTRAINT chat_file_links_chat_id_file_id_key UNIQUE (chat_id, file_id);
+
 ALTER TABLE ONLY chat_files
     ADD CONSTRAINT chat_files_pkey PRIMARY KEY (id);
 
@@ -3545,6 +3574,12 @@ ALTER TABLE ONLY usage_events_daily
 ALTER TABLE ONLY usage_events
     ADD CONSTRAINT usage_events_pkey PRIMARY KEY (id);
 
+ALTER TABLE ONLY user_chat_provider_keys
+    ADD CONSTRAINT user_chat_provider_keys_pkey PRIMARY KEY (id);
+
+ALTER TABLE ONLY user_chat_provider_keys
+    ADD CONSTRAINT user_chat_provider_keys_user_id_chat_provider_id_key UNIQUE (user_id, chat_provider_id);
+
 ALTER TABLE ONLY user_configs
     ADD CONSTRAINT user_configs_pkey PRIMARY KEY (user_id, key);
 
@@ -3707,6 +3742,8 @@ CREATE INDEX idx_audit_logs_time_desc ON audit_logs USING btree ("time" DESC);
 
 CREATE INDEX idx_chat_diff_statuses_stale_at ON chat_diff_statuses USING btree (stale_at);
 
+CREATE INDEX idx_chat_file_links_chat_id ON chat_file_links USING btree (chat_id);
+
 CREATE INDEX idx_chat_files_org ON chat_files USING btree (organization_id);
 
 CREATE INDEX idx_chat_files_owner ON chat_files USING btree (owner_id);
@@ -4009,6 +4046,12 @@ ALTER TABLE ONLY api_keys
 ALTER TABLE ONLY chat_diff_statuses
     ADD CONSTRAINT chat_diff_statuses_chat_id_fkey FOREIGN KEY (chat_id) REFERENCES chats(id) ON DELETE CASCADE;
 
+ALTER TABLE ONLY chat_file_links
+    ADD CONSTRAINT chat_file_links_chat_id_fkey FOREIGN KEY (chat_id) REFERENCES chats(id) ON DELETE CASCADE;
+
+ALTER TABLE ONLY chat_file_links
+    ADD CONSTRAINT chat_file_links_file_id_fkey FOREIGN KEY (file_id) REFERENCES chat_files(id) ON DELETE CASCADE;
+
 ALTER TABLE ONLY chat_files
     ADD CONSTRAINT chat_files_organization_id_fkey FOREIGN KEY (organization_id) REFERENCES organizations(id) ON DELETE CASCADE;
 
@@ -4255,6 +4298,15 @@ ALTER TABLE ONLY templates
 ALTER TABLE ONLY templates
     ADD CONSTRAINT templates_organization_id_fkey FOREIGN KEY (organization_id) REFERENCES organizations(id) ON DELETE CASCADE;
 
+ALTER TABLE ONLY user_chat_provider_keys
+    ADD CONSTRAINT user_chat_provider_keys_api_key_key_id_fkey FOREIGN KEY (api_key_key_id) REFERENCES dbcrypt_keys(active_key_digest);
+
+ALTER TABLE ONLY user_chat_provider_keys
+    ADD CONSTRAINT user_chat_provider_keys_chat_provider_id_fkey FOREIGN KEY (chat_provider_id) REFERENCES chat_providers(id) ON DELETE CASCADE;
+
+ALTER TABLE ONLY user_chat_provider_keys
+    ADD CONSTRAINT user_chat_provider_keys_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
+
 ALTER TABLE ONLY user_configs
     ADD CONSTRAINT user_configs_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
 
@@ -4273,6 +4325,9 @@ ALTER TABLE ONLY user_links
 ALTER TABLE ONLY user_secrets
     ADD CONSTRAINT user_secrets_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
 
+ALTER TABLE ONLY user_secrets
+    ADD CONSTRAINT user_secrets_value_key_id_fkey FOREIGN KEY (value_key_id) REFERENCES dbcrypt_keys(active_key_digest);
+
 ALTER TABLE ONLY user_status_changes
     ADD CONSTRAINT user_status_changes_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id);
 

coderd/database/foreign_key_constraint.go

@@ -10,6 +10,8 @@ const (
 	ForeignKeyAibridgeInterceptionsInitiatorID                    ForeignKeyConstraint = "aibridge_interceptions_initiator_id_fkey"                        // ALTER TABLE ONLY aibridge_interceptions ADD CONSTRAINT aibridge_interceptions_initiator_id_fkey FOREIGN KEY (initiator_id) REFERENCES users(id);
 	ForeignKeyAPIKeysUserIDUUID                                   ForeignKeyConstraint = "api_keys_user_id_uuid_fkey"                                      // ALTER TABLE ONLY api_keys ADD CONSTRAINT api_keys_user_id_uuid_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
 	ForeignKeyChatDiffStatusesChatID                              ForeignKeyConstraint = "chat_diff_statuses_chat_id_fkey"                                 // ALTER TABLE ONLY chat_diff_statuses ADD CONSTRAINT chat_diff_statuses_chat_id_fkey FOREIGN KEY (chat_id) REFERENCES chats(id) ON DELETE CASCADE;
+	ForeignKeyChatFileLinksChatID                                 ForeignKeyConstraint = "chat_file_links_chat_id_fkey"                                    // ALTER TABLE ONLY chat_file_links ADD CONSTRAINT chat_file_links_chat_id_fkey FOREIGN KEY (chat_id) REFERENCES chats(id) ON DELETE CASCADE;
+	ForeignKeyChatFileLinksFileID                                 ForeignKeyConstraint = "chat_file_links_file_id_fkey"                                    // ALTER TABLE ONLY chat_file_links ADD CONSTRAINT chat_file_links_file_id_fkey FOREIGN KEY (file_id) REFERENCES chat_files(id) ON DELETE CASCADE;
 	ForeignKeyChatFilesOrganizationID                             ForeignKeyConstraint = "chat_files_organization_id_fkey"                                 // ALTER TABLE ONLY chat_files ADD CONSTRAINT chat_files_organization_id_fkey FOREIGN KEY (organization_id) REFERENCES organizations(id) ON DELETE CASCADE;
 	ForeignKeyChatFilesOwnerID                                    ForeignKeyConstraint = "chat_files_owner_id_fkey"                                        // ALTER TABLE ONLY chat_files ADD CONSTRAINT chat_files_owner_id_fkey FOREIGN KEY (owner_id) REFERENCES users(id) ON DELETE CASCADE;
 	ForeignKeyChatMessagesChatID                                  ForeignKeyConstraint = "chat_messages_chat_id_fkey"                                      // ALTER TABLE ONLY chat_messages ADD CONSTRAINT chat_messages_chat_id_fkey FOREIGN KEY (chat_id) REFERENCES chats(id) ON DELETE CASCADE;
@@ -92,12 +94,16 @@ const (
 	ForeignKeyTemplateVersionsTemplateID                          ForeignKeyConstraint = "template_versions_template_id_fkey"                              // ALTER TABLE ONLY template_versions ADD CONSTRAINT template_versions_template_id_fkey FOREIGN KEY (template_id) REFERENCES templates(id) ON DELETE CASCADE;
 	ForeignKeyTemplatesCreatedBy                                  ForeignKeyConstraint = "templates_created_by_fkey"                                       // ALTER TABLE ONLY templates ADD CONSTRAINT templates_created_by_fkey FOREIGN KEY (created_by) REFERENCES users(id) ON DELETE RESTRICT;
 	ForeignKeyTemplatesOrganizationID                             ForeignKeyConstraint = "templates_organization_id_fkey"                                  // ALTER TABLE ONLY templates ADD CONSTRAINT templates_organization_id_fkey FOREIGN KEY (organization_id) REFERENCES organizations(id) ON DELETE CASCADE;
+	ForeignKeyUserChatProviderKeysAPIKeyKeyID                     ForeignKeyConstraint = "user_chat_provider_keys_api_key_key_id_fkey"                     // ALTER TABLE ONLY user_chat_provider_keys ADD CONSTRAINT user_chat_provider_keys_api_key_key_id_fkey FOREIGN KEY (api_key_key_id) REFERENCES dbcrypt_keys(active_key_digest);
+	ForeignKeyUserChatProviderKeysChatProviderID                  ForeignKeyConstraint = "user_chat_provider_keys_chat_provider_id_fkey"                   // ALTER TABLE ONLY user_chat_provider_keys ADD CONSTRAINT user_chat_provider_keys_chat_provider_id_fkey FOREIGN KEY (chat_provider_id) REFERENCES chat_providers(id) ON DELETE CASCADE;
+	ForeignKeyUserChatProviderKeysUserID                          ForeignKeyConstraint = "user_chat_provider_keys_user_id_fkey"                            // ALTER TABLE ONLY user_chat_provider_keys ADD CONSTRAINT user_chat_provider_keys_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
 	ForeignKeyUserConfigsUserID                                   ForeignKeyConstraint = "user_configs_user_id_fkey"                                       // ALTER TABLE ONLY user_configs ADD CONSTRAINT user_configs_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
 	ForeignKeyUserDeletedUserID                                   ForeignKeyConstraint = "user_deleted_user_id_fkey"                                       // ALTER TABLE ONLY user_deleted ADD CONSTRAINT user_deleted_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id);
 	ForeignKeyUserLinksOauthAccessTokenKeyID                      ForeignKeyConstraint = "user_links_oauth_access_token_key_id_fkey"                       // ALTER TABLE ONLY user_links ADD CONSTRAINT user_links_oauth_access_token_key_id_fkey FOREIGN KEY (oauth_access_token_key_id) REFERENCES dbcrypt_keys(active_key_digest);
 	ForeignKeyUserLinksOauthRefreshTokenKeyID                     ForeignKeyConstraint = "user_links_oauth_refresh_token_key_id_fkey"                      // ALTER TABLE ONLY user_links ADD CONSTRAINT user_links_oauth_refresh_token_key_id_fkey FOREIGN KEY (oauth_refresh_token_key_id) REFERENCES dbcrypt_keys(active_key_digest);
 	ForeignKeyUserLinksUserID                                     ForeignKeyConstraint = "user_links_user_id_fkey"                                         // ALTER TABLE ONLY user_links ADD CONSTRAINT user_links_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
 	ForeignKeyUserSecretsUserID                                   ForeignKeyConstraint = "user_secrets_user_id_fkey"                                       // ALTER TABLE ONLY user_secrets ADD CONSTRAINT user_secrets_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
+	ForeignKeyUserSecretsValueKeyID                               ForeignKeyConstraint = "user_secrets_value_key_id_fkey"                                  // ALTER TABLE ONLY user_secrets ADD CONSTRAINT user_secrets_value_key_id_fkey FOREIGN KEY (value_key_id) REFERENCES dbcrypt_keys(active_key_digest);
 	ForeignKeyUserStatusChangesUserID                             ForeignKeyConstraint = "user_status_changes_user_id_fkey"                                // ALTER TABLE ONLY user_status_changes ADD CONSTRAINT user_status_changes_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id);
 	ForeignKeyWebpushSubscriptionsUserID                          ForeignKeyConstraint = "webpush_subscriptions_user_id_fkey"                              // ALTER TABLE ONLY webpush_subscriptions ADD CONSTRAINT webpush_subscriptions_user_id_fkey FOREIGN KEY (user_id) REFERENCES users(id) ON DELETE CASCADE;
 	ForeignKeyWorkspaceAgentDevcontainersSubagentID               ForeignKeyConstraint = "workspace_agent_devcontainers_subagent_id_fkey"                  // ALTER TABLE ONLY workspace_agent_devcontainers ADD CONSTRAINT workspace_agent_devcontainers_subagent_id_fkey FOREIGN KEY (subagent_id) REFERENCES workspace_agents(id) ON DELETE CASCADE;

coderd/database/migrations/000458_aibridge_provider_name.down.sql

@@ -0,0 +1 @@
+ALTER TABLE aibridge_interceptions DROP COLUMN provider_name;

coderd/database/migrations/000458_aibridge_provider_name.up.sql

@@ -0,0 +1,6 @@
+ALTER TABLE aibridge_interceptions ADD COLUMN provider_name TEXT NOT NULL DEFAULT '';
+
+COMMENT ON COLUMN aibridge_interceptions.provider_name IS 'The provider instance name which may differ from provider when multiple instances of the same provider type exist.';
+
+-- Backfill existing records with the provider type as the provider name.
+UPDATE aibridge_interceptions SET provider_name = provider WHERE provider_name = '';

coderd/database/migrations/000459_provider_key_policy.down.sql

@@ -0,0 +1,8 @@
+DROP TABLE IF EXISTS user_chat_provider_keys;
+
+ALTER TABLE chat_providers DROP CONSTRAINT IF EXISTS valid_credential_policy;
+
+ALTER TABLE chat_providers
+    DROP COLUMN IF EXISTS central_api_key_enabled,
+    DROP COLUMN IF EXISTS allow_user_api_key,
+    DROP COLUMN IF EXISTS allow_central_api_key_fallback;

coderd/database/migrations/000459_provider_key_policy.up.sql

@@ -0,0 +1,24 @@
+ALTER TABLE chat_providers
+    ADD COLUMN central_api_key_enabled BOOLEAN NOT NULL DEFAULT TRUE,
+    ADD COLUMN allow_user_api_key BOOLEAN NOT NULL DEFAULT FALSE,
+    ADD COLUMN allow_central_api_key_fallback BOOLEAN NOT NULL DEFAULT FALSE;
+
+ALTER TABLE chat_providers
+    ADD CONSTRAINT valid_credential_policy CHECK (
+        (central_api_key_enabled OR allow_user_api_key) AND
+        (
+            NOT allow_central_api_key_fallback OR
+            (central_api_key_enabled AND allow_user_api_key)
+        )
+    );
+
+CREATE TABLE user_chat_provider_keys (
+    id               UUID        PRIMARY KEY DEFAULT gen_random_uuid(),
+    user_id          UUID        NOT NULL REFERENCES users(id) ON DELETE CASCADE,
+    chat_provider_id UUID        NOT NULL REFERENCES chat_providers(id) ON DELETE CASCADE,
+    api_key          TEXT        NOT NULL CHECK (api_key != ''),
+    api_key_key_id   TEXT        REFERENCES dbcrypt_keys(active_key_digest),
+    created_at       TIMESTAMPTZ NOT NULL DEFAULT NOW(),
+    updated_at       TIMESTAMPTZ NOT NULL DEFAULT NOW(),
+    UNIQUE (user_id, chat_provider_id)
+);

coderd/database/migrations/000460_user_secrets_value_key_id.down.sql

@@ -0,0 +1,3 @@
+ALTER TABLE user_secrets
+    DROP CONSTRAINT user_secrets_value_key_id_fkey,
+    DROP COLUMN value_key_id;

coderd/database/migrations/000460_user_secrets_value_key_id.up.sql

@@ -0,0 +1,5 @@
+ALTER TABLE user_secrets
+    ADD COLUMN value_key_id TEXT;
+
+ALTER TABLE ONLY user_secrets
+    ADD CONSTRAINT user_secrets_value_key_id_fkey FOREIGN KEY (value_key_id) REFERENCES dbcrypt_keys(active_key_digest);

coderd/database/migrations/000461_aibridge_cache_token_columns.down.sql

@@ -0,0 +1,3 @@
+ALTER TABLE aibridge_token_usages
+    DROP COLUMN cache_read_input_tokens,
+    DROP COLUMN cache_write_input_tokens;

coderd/database/migrations/000461_aibridge_cache_token_columns.up.sql

@@ -0,0 +1,26 @@
+ALTER TABLE aibridge_token_usages
+    ADD COLUMN cache_read_input_tokens BIGINT NOT NULL DEFAULT 0,
+    ADD COLUMN cache_write_input_tokens BIGINT NOT NULL DEFAULT 0;
+
+-- Backfill from metadata JSONB. Old rows stored cache tokens under
+-- provider-specific keys; new rows use the dedicated columns above.
+UPDATE aibridge_token_usages
+SET
+
+    -- Cache-read metadata keys by provider:
+    --   Anthropic (/v1/messages):          "cache_read_input"
+    --   OpenAI    (/v1/responses):         "input_cached"
+    --   OpenAI    (/v1/chat/completions):  "prompt_cached"
+    cache_read_input_tokens = GREATEST(
+        COALESCE((metadata->>'cache_read_input')::bigint, 0),
+        COALESCE((metadata->>'input_cached')::bigint, 0),
+        COALESCE((metadata->>'prompt_cached')::bigint, 0)
+    ),
+
+	-- Cache-write metadata keys by provider:
+	--   Anthropic (/v1/messages):          "cache_creation_input"
+	--   OpenAI does not report cache-write tokens.
+    cache_write_input_tokens = COALESCE((metadata->>'cache_creation_input')::bigint, 0)
+WHERE metadata IS NOT NULL
+  AND cache_read_input_tokens = 0
+  AND cache_write_input_tokens = 0;

coderd/database/migrations/000462_chat_file_links.down.sql

@@ -0,0 +1,9 @@
+ALTER TABLE chats ADD COLUMN file_ids uuid[] DEFAULT '{}'::uuid[] NOT NULL;
+
+UPDATE chats SET file_ids = (
+    SELECT COALESCE(array_agg(cfl.file_id), '{}')
+    FROM chat_file_links cfl
+    WHERE cfl.chat_id = chats.id
+);
+
+DROP TABLE chat_file_links;

coderd/database/migrations/000462_chat_file_links.up.sql

@@ -0,0 +1,17 @@
+CREATE TABLE chat_file_links (
+    chat_id uuid NOT NULL,
+    file_id uuid NOT NULL,
+    UNIQUE (chat_id, file_id)
+);
+
+CREATE INDEX idx_chat_file_links_chat_id ON chat_file_links (chat_id);
+
+ALTER TABLE chat_file_links
+    ADD CONSTRAINT chat_file_links_chat_id_fkey
+    FOREIGN KEY (chat_id) REFERENCES chats(id) ON DELETE CASCADE;
+
+ALTER TABLE chat_file_links
+    ADD CONSTRAINT chat_file_links_file_id_fkey
+    FOREIGN KEY (file_id) REFERENCES chat_files(id) ON DELETE CASCADE;
+
+ALTER TABLE chats DROP COLUMN IF EXISTS file_ids;

coderd/database/migrations/testdata/fixtures/000459_provider_key_policy.up.sql

@@ -0,0 +1,16 @@
+INSERT INTO user_chat_provider_keys (
+    user_id,
+    chat_provider_id,
+    api_key,
+    created_at,
+    updated_at
+)
+SELECT
+    id,
+    '0a8b2f84-b5a8-4c44-8c9f-e58c44a534a7',
+    'fixture-test-key',
+    '2025-01-01 00:00:00+00',
+    '2025-01-01 00:00:00+00'
+FROM users
+ORDER BY created_at, id
+LIMIT 1;

coderd/database/migrations/testdata/fixtures/000462_chat_file_links.up.sql

@@ -0,0 +1,5 @@
+INSERT INTO chat_file_links (chat_id, file_id)
+VALUES (
+    '72c0438a-18eb-4688-ab80-e4c6a126ef96',
+    '00000000-0000-0000-0000-000000000099'
+);

coderd/database/modelmethods.go

@@ -10,6 +10,7 @@ import (
 	"time"
 
 	"github.com/google/uuid"
+	"github.com/sqlc-dev/pqtype"
 	"golang.org/x/exp/maps"
 	"golang.org/x/oauth2"
 	"golang.org/x/xerrors"
@@ -186,6 +187,10 @@ func (c ChatFile) RBACObject() rbac.Object {
 	return rbac.ResourceChat.WithID(c.ID).WithOwner(c.OwnerID.String()).InOrg(c.OrganizationID)
 }
 
+func (c GetChatFileMetadataByChatIDRow) RBACObject() rbac.Object {
+	return rbac.ResourceChat.WithID(c.ID).WithOwner(c.OwnerID.String()).InOrg(c.OrganizationID)
+}
+
 func (s APIKeyScope) ToRBAC() rbac.ScopeName {
 	switch s {
 	case ApiKeyScopeCoderAll:
@@ -923,3 +928,28 @@ func WorkspaceIdentityFromWorkspace(w Workspace) WorkspaceIdentity {
 func (r GetWorkspaceAgentAndWorkspaceByIDRow) RBACObject() rbac.Object {
 	return r.WorkspaceTable.RBACObject()
 }
+
+// UpsertConnectionLogParams contains the parameters for upserting a
+// connection log entry. This struct is hand-maintained (not generated
+// by sqlc) because the single-row UpsertConnectionLog query was
+// removed in favor of BatchUpsertConnectionLogs, but the struct is
+// still used as the canonical connection log event type throughout
+// the codebase.
+type UpsertConnectionLogParams struct {
+	ID               uuid.UUID        `db:"id" json:"id"`
+	OrganizationID   uuid.UUID        `db:"organization_id" json:"organization_id"`
+	WorkspaceOwnerID uuid.UUID        `db:"workspace_owner_id" json:"workspace_owner_id"`
+	WorkspaceID      uuid.UUID        `db:"workspace_id" json:"workspace_id"`
+	WorkspaceName    string           `db:"workspace_name" json:"workspace_name"`
+	AgentName        string           `db:"agent_name" json:"agent_name"`
+	Type             ConnectionType   `db:"type" json:"type"`
+	Code             sql.NullInt32    `db:"code" json:"code"`
+	IP               pqtype.Inet      `db:"ip" json:"ip"`
+	UserAgent        sql.NullString   `db:"user_agent" json:"user_agent"`
+	UserID           uuid.NullUUID    `db:"user_id" json:"user_id"`
+	SlugOrPort       sql.NullString   `db:"slug_or_port" json:"slug_or_port"`
+	ConnectionID     uuid.NullUUID    `db:"connection_id" json:"connection_id"`
+	DisconnectReason sql.NullString   `db:"disconnect_reason" json:"disconnect_reason"`
+	Time             time.Time        `db:"time" json:"time"`
+	ConnectionStatus ConnectionStatus `db:"connection_status" json:"connection_status"`
+}

coderd/database/modelqueries.go

@@ -584,6 +584,7 @@ func (q *sqlQuerier) CountAuthorizedAuditLogs(ctx context.Context, arg CountAudi
 		arg.DateTo,
 		arg.BuildReason,
 		arg.RequestID,
+		arg.CountCap,
 	)
 	if err != nil {
 		return 0, err
@@ -720,6 +721,7 @@ func (q *sqlQuerier) CountAuthorizedConnectionLogs(ctx context.Context, arg Coun
 		arg.WorkspaceID,
 		arg.ConnectionID,
 		arg.Status,
+		arg.CountCap,
 	)
 	if err != nil {
 		return 0, err
@@ -865,6 +867,7 @@ func (q *sqlQuerier) ListAuthorizedAIBridgeInterceptions(ctx context.Context, ar
 			&i.AIBridgeInterception.ThreadRootID,
 			&i.AIBridgeInterception.ClientSessionID,
 			&i.AIBridgeInterception.SessionID,
+			&i.AIBridgeInterception.ProviderName,
 			&i.VisibleUser.ID,
 			&i.VisibleUser.Username,
 			&i.VisibleUser.Name,
@@ -1028,6 +1031,8 @@ func (q *sqlQuerier) ListAuthorizedAIBridgeSessions(ctx context.Context, arg Lis
 			&i.Threads,
 			&i.InputTokens,
 			&i.OutputTokens,
+			&i.CacheReadInputTokens,
+			&i.CacheWriteInputTokens,
 			&i.LastPrompt,
 		); err != nil {
 			return nil, err
@@ -1125,6 +1130,7 @@ func (q *sqlQuerier) ListAuthorizedAIBridgeSessionThreads(ctx context.Context, a
 			&i.AIBridgeInterception.ThreadRootID,
 			&i.AIBridgeInterception.ClientSessionID,
 			&i.AIBridgeInterception.SessionID,
+			&i.AIBridgeInterception.ProviderName,
 		); err != nil {
 			return nil, err
 		}

coderd/database/modelqueries_internal_test.go

@@ -145,5 +145,13 @@ func extractWhereClause(query string) string {
 	// Remove SQL comments
 	whereClause = regexp.MustCompile(`(?m)--.*$`).ReplaceAllString(whereClause, "")
 
+	// Normalize indentation so subquery wrapping doesn't cause
+	// mismatches.
+	lines := strings.Split(whereClause, "\n")
+	for i, line := range lines {
+		lines[i] = strings.TrimLeft(line, " \t")
+	}
+	whereClause = strings.Join(lines, "\n")
+
 	return strings.TrimSpace(whereClause)
 }

coderd/database/models.go

@@ -4038,6 +4038,8 @@ type AIBridgeInterception struct {
 	ClientSessionID sql.NullString `db:"client_session_id" json:"client_session_id"`
 	// Groups related interceptions into a logical session. Determined by a priority chain: (1) client_session_id — an explicit session identifier supplied by the calling client (e.g. Claude Code); (2) thread_root_id — the root of an agentic thread detected by Bridge through tool-call correlation, used when the client does not supply its own session ID; (3) id — the interception's own ID, used as a last resort so every interception belongs to exactly one session even if it is standalone. This is a generated column stored on disk so it can be indexed and joined without recomputing the COALESCE on every query.
 	SessionID string `db:"session_id" json:"session_id"`
+	// The provider instance name which may differ from provider when multiple instances of the same provider type exist.
+	ProviderName string `db:"provider_name" json:"provider_name"`
 }
 
 // Audit log of model thinking in intercepted requests in AI Bridge
@@ -4053,11 +4055,13 @@ type AIBridgeTokenUsage struct {
 	ID             uuid.UUID `db:"id" json:"id"`
 	InterceptionID uuid.UUID `db:"interception_id" json:"interception_id"`
 	// The ID for the response in which the tokens were used, produced by the provider.
-	ProviderResponseID string                `db:"provider_response_id" json:"provider_response_id"`
-	InputTokens        int64                 `db:"input_tokens" json:"input_tokens"`
-	OutputTokens       int64                 `db:"output_tokens" json:"output_tokens"`
-	Metadata           pqtype.NullRawMessage `db:"metadata" json:"metadata"`
-	CreatedAt          time.Time             `db:"created_at" json:"created_at"`
+	ProviderResponseID    string                `db:"provider_response_id" json:"provider_response_id"`
+	InputTokens           int64                 `db:"input_tokens" json:"input_tokens"`
+	OutputTokens          int64                 `db:"output_tokens" json:"output_tokens"`
+	Metadata              pqtype.NullRawMessage `db:"metadata" json:"metadata"`
+	CreatedAt             time.Time             `db:"created_at" json:"created_at"`
+	CacheReadInputTokens  int64                 `db:"cache_read_input_tokens" json:"cache_read_input_tokens"`
+	CacheWriteInputTokens int64                 `db:"cache_write_input_tokens" json:"cache_write_input_tokens"`
 }
 
 // Audit log of tool calls in intercepted requests in AI Bridge
@@ -4214,6 +4218,11 @@ type ChatFile struct {
 	Data           []byte    `db:"data" json:"data"`
 }
 
+type ChatFileLink struct {
+	ChatID uuid.UUID `db:"chat_id" json:"chat_id"`
+	FileID uuid.UUID `db:"file_id" json:"file_id"`
+}
+
 type ChatMessage struct {
 	ID                  int64                 `db:"id" json:"id"`
 	ChatID              uuid.UUID             `db:"chat_id" json:"chat_id"`
@@ -4262,12 +4271,15 @@ type ChatProvider struct {
 	DisplayName string    `db:"display_name" json:"display_name"`
 	APIKey      string    `db:"api_key" json:"api_key"`
 	// The ID of the key used to encrypt the provider API key. If this is NULL, the API key is not encrypted
-	ApiKeyKeyID sql.NullString `db:"api_key_key_id" json:"api_key_key_id"`
-	CreatedBy   uuid.NullUUID  `db:"created_by" json:"created_by"`
-	Enabled     bool           `db:"enabled" json:"enabled"`
-	CreatedAt   time.Time      `db:"created_at" json:"created_at"`
-	UpdatedAt   time.Time      `db:"updated_at" json:"updated_at"`
-	BaseUrl     string         `db:"base_url" json:"base_url"`
+	ApiKeyKeyID                sql.NullString `db:"api_key_key_id" json:"api_key_key_id"`
+	CreatedBy                  uuid.NullUUID  `db:"created_by" json:"created_by"`
+	Enabled                    bool           `db:"enabled" json:"enabled"`
+	CreatedAt                  time.Time      `db:"created_at" json:"created_at"`
+	UpdatedAt                  time.Time      `db:"updated_at" json:"updated_at"`
+	BaseUrl                    string         `db:"base_url" json:"base_url"`
+	CentralApiKeyEnabled       bool           `db:"central_api_key_enabled" json:"central_api_key_enabled"`
+	AllowUserApiKey            bool           `db:"allow_user_api_key" json:"allow_user_api_key"`
+	AllowCentralApiKeyFallback bool           `db:"allow_central_api_key_fallback" json:"allow_central_api_key_fallback"`
 }
 
 type ChatQueuedMessage struct {
@@ -5220,6 +5232,16 @@ type User struct {
 	ChatSpendLimitMicros sql.NullInt64 `db:"chat_spend_limit_micros" json:"chat_spend_limit_micros"`
 }
 
+type UserChatProviderKey struct {
+	ID             uuid.UUID      `db:"id" json:"id"`
+	UserID         uuid.UUID      `db:"user_id" json:"user_id"`
+	ChatProviderID uuid.UUID      `db:"chat_provider_id" json:"chat_provider_id"`
+	APIKey         string         `db:"api_key" json:"api_key"`
+	ApiKeyKeyID    sql.NullString `db:"api_key_key_id" json:"api_key_key_id"`
+	CreatedAt      time.Time      `db:"created_at" json:"created_at"`
+	UpdatedAt      time.Time      `db:"updated_at" json:"updated_at"`
+}
+
 type UserConfig struct {
 	UserID uuid.UUID `db:"user_id" json:"user_id"`
 	Key    string    `db:"key" json:"key"`
@@ -5249,15 +5271,16 @@ type UserLink struct {
 }
 
 type UserSecret struct {
-	ID          uuid.UUID `db:"id" json:"id"`
-	UserID      uuid.UUID `db:"user_id" json:"user_id"`
-	Name        string    `db:"name" json:"name"`
-	Description string    `db:"description" json:"description"`
-	Value       string    `db:"value" json:"value"`
-	EnvName     string    `db:"env_name" json:"env_name"`
-	FilePath    string    `db:"file_path" json:"file_path"`
-	CreatedAt   time.Time `db:"created_at" json:"created_at"`
-	UpdatedAt   time.Time `db:"updated_at" json:"updated_at"`
+	ID          uuid.UUID      `db:"id" json:"id"`
+	UserID      uuid.UUID      `db:"user_id" json:"user_id"`
+	Name        string         `db:"name" json:"name"`
+	Description string         `db:"description" json:"description"`
+	Value       string         `db:"value" json:"value"`
+	EnvName     string         `db:"env_name" json:"env_name"`
+	FilePath    string         `db:"file_path" json:"file_path"`
+	CreatedAt   time.Time      `db:"created_at" json:"created_at"`
+	UpdatedAt   time.Time      `db:"updated_at" json:"updated_at"`
+	ValueKeyID  sql.NullString `db:"value_key_id" json:"value_key_id"`
 }
 
 // Tracks the history of user status changes

coderd/database/querier.go

@@ -54,7 +54,7 @@ type sqlcQuerier interface {
 	ActivityBumpWorkspace(ctx context.Context, arg ActivityBumpWorkspaceParams) error
 	// AllUserIDs returns all UserIDs regardless of user status or deletion.
 	AllUserIDs(ctx context.Context, includeSystem bool) ([]uuid.UUID, error)
-	ArchiveChatByID(ctx context.Context, id uuid.UUID) error
+	ArchiveChatByID(ctx context.Context, id uuid.UUID) ([]Chat, error)
 	// Archiving templates is a soft delete action, so is reversible.
 	// Archiving prevents the version from being used and discovered
 	// by listing.
@@ -65,6 +65,7 @@ type sqlcQuerier interface {
 	BatchUpdateWorkspaceAgentMetadata(ctx context.Context, arg BatchUpdateWorkspaceAgentMetadataParams) error
 	BatchUpdateWorkspaceLastUsedAt(ctx context.Context, arg BatchUpdateWorkspaceLastUsedAtParams) error
 	BatchUpdateWorkspaceNextStartAt(ctx context.Context, arg BatchUpdateWorkspaceNextStartAtParams) error
+	BatchUpsertConnectionLogs(ctx context.Context, arg BatchUpsertConnectionLogsParams) error
 	BulkMarkNotificationMessagesFailed(ctx context.Context, arg BulkMarkNotificationMessagesFailedParams) (int64, error)
 	BulkMarkNotificationMessagesSent(ctx context.Context, arg BulkMarkNotificationMessagesSentParams) (int64, error)
 	// Calculates the telemetry summary for a given provider, model, and client
@@ -150,6 +151,7 @@ type sqlcQuerier interface {
 	DeleteTailnetTunnel(ctx context.Context, arg DeleteTailnetTunnelParams) (DeleteTailnetTunnelRow, error)
 	DeleteTask(ctx context.Context, arg DeleteTaskParams) (uuid.UUID, error)
 	DeleteUserChatCompactionThreshold(ctx context.Context, arg DeleteUserChatCompactionThresholdParams) error
+	DeleteUserChatProviderKey(ctx context.Context, arg DeleteUserChatProviderKeyParams) error
 	DeleteUserSecret(ctx context.Context, id uuid.UUID) error
 	DeleteWebpushSubscriptionByUserIDAndEndpoint(ctx context.Context, arg DeleteWebpushSubscriptionByUserIDAndEndpointParams) error
 	DeleteWebpushSubscriptions(ctx context.Context, ids []uuid.UUID) error
@@ -242,6 +244,10 @@ type sqlcQuerier interface {
 	GetChatDiffStatusByChatID(ctx context.Context, chatID uuid.UUID) (ChatDiffStatus, error)
 	GetChatDiffStatusesByChatIDs(ctx context.Context, chatIds []uuid.UUID) ([]ChatDiffStatus, error)
 	GetChatFileByID(ctx context.Context, id uuid.UUID) (ChatFile, error)
+	// GetChatFileMetadataByChatID returns lightweight file metadata for
+	// all files linked to a chat. The data column is excluded to avoid
+	// loading file content.
+	GetChatFileMetadataByChatID(ctx context.Context, chatID uuid.UUID) ([]GetChatFileMetadataByChatIDRow, error)
 	GetChatFilesByIDs(ctx context.Context, ids []uuid.UUID) ([]ChatFile, error)
 	// GetChatIncludeDefaultSystemPrompt preserves the legacy default
 	// for deployments created before the explicit include-default toggle.
@@ -476,8 +482,8 @@ type sqlcQuerier interface {
 	// Used for recovery after coderd crashes or long hangs.
 	GetStaleChats(ctx context.Context, staleThreshold time.Time) ([]Chat, error)
 	GetTailnetPeers(ctx context.Context, id uuid.UUID) ([]TailnetPeer, error)
-	GetTailnetTunnelPeerBindings(ctx context.Context, srcID uuid.UUID) ([]GetTailnetTunnelPeerBindingsRow, error)
-	GetTailnetTunnelPeerIDs(ctx context.Context, srcID uuid.UUID) ([]GetTailnetTunnelPeerIDsRow, error)
+	GetTailnetTunnelPeerBindingsBatch(ctx context.Context, ids []uuid.UUID) ([]GetTailnetTunnelPeerBindingsBatchRow, error)
+	GetTailnetTunnelPeerIDsBatch(ctx context.Context, ids []uuid.UUID) ([]GetTailnetTunnelPeerIDsBatchRow, error)
 	GetTaskByID(ctx context.Context, id uuid.UUID) (Task, error)
 	GetTaskByOwnerIDAndName(ctx context.Context, arg GetTaskByOwnerIDAndNameParams) (Task, error)
 	GetTaskByWorkspaceID(ctx context.Context, workspaceID uuid.UUID) (Task, error)
@@ -577,6 +583,7 @@ type sqlcQuerier interface {
 	GetUserByID(ctx context.Context, id uuid.UUID) (User, error)
 	GetUserChatCompactionThreshold(ctx context.Context, arg GetUserChatCompactionThresholdParams) (string, error)
 	GetUserChatCustomPrompt(ctx context.Context, userID uuid.UUID) (string, error)
+	GetUserChatProviderKeys(ctx context.Context, userID uuid.UUID) ([]UserChatProviderKey, error)
 	GetUserChatSpendInPeriod(ctx context.Context, arg GetUserChatSpendInPeriodParams) (int64, error)
 	GetUserCount(ctx context.Context, includeSystem bool) (int64, error)
 	// Returns the minimum (most restrictive) group limit for a user.
@@ -775,6 +782,15 @@ type sqlcQuerier interface {
 	InsertWorkspaceProxy(ctx context.Context, arg InsertWorkspaceProxyParams) (WorkspaceProxy, error)
 	InsertWorkspaceResource(ctx context.Context, arg InsertWorkspaceResourceParams) (WorkspaceResource, error)
 	InsertWorkspaceResourceMetadata(ctx context.Context, arg InsertWorkspaceResourceMetadataParams) ([]WorkspaceResourceMetadatum, error)
+	// LinkChatFiles inserts file associations into the chat_file_links
+	// join table with deduplication (ON CONFLICT DO NOTHING). The INSERT
+	// is conditional: it only proceeds when the total number of links
+	// (existing + genuinely new) does not exceed max_file_links. Returns
+	// the number of genuinely new file IDs that were NOT inserted due to
+	// the cap. A return value of 0 means all files were linked (or were
+	// already linked). A positive value means the cap blocked that many
+	// new links.
+	LinkChatFiles(ctx context.Context, arg LinkChatFilesParams) (int32, error)
 	ListAIBridgeClients(ctx context.Context, arg ListAIBridgeClientsParams) ([]string, error)
 	ListAIBridgeInterceptions(ctx context.Context, arg ListAIBridgeInterceptionsParams) ([]ListAIBridgeInterceptionsRow, error)
 	// Finds all unique AI Bridge interception telemetry summaries combinations
@@ -844,7 +860,7 @@ type sqlcQuerier interface {
 	// This must be called from within a transaction. The lock will be automatically
 	// released when the transaction ends.
 	TryAcquireLock(ctx context.Context, pgTryAdvisoryXactLock int64) (bool, error)
-	UnarchiveChatByID(ctx context.Context, id uuid.UUID) error
+	UnarchiveChatByID(ctx context.Context, id uuid.UUID) ([]Chat, error)
 	// This will always work regardless of the current state of the template version.
 	UnarchiveTemplateVersion(ctx context.Context, arg UnarchiveTemplateVersionParams) error
 	UnfavoriteWorkspace(ctx context.Context, id uuid.UUID) error
@@ -854,9 +870,11 @@ type sqlcQuerier interface {
 	UpdateAPIKeyByID(ctx context.Context, arg UpdateAPIKeyByIDParams) error
 	UpdateChatBuildAgentBinding(ctx context.Context, arg UpdateChatBuildAgentBindingParams) (Chat, error)
 	UpdateChatByID(ctx context.Context, arg UpdateChatByIDParams) (Chat, error)
-	// Bumps the heartbeat timestamp for a running chat so that other
-	// replicas know the worker is still alive.
-	UpdateChatHeartbeat(ctx context.Context, arg UpdateChatHeartbeatParams) (int64, error)
+	// Bumps the heartbeat timestamp for the given set of chat IDs,
+	// provided they are still running and owned by the specified
+	// worker. Returns the IDs that were actually updated so the
+	// caller can detect stolen or completed chats via set-difference.
+	UpdateChatHeartbeats(ctx context.Context, arg UpdateChatHeartbeatsParams) ([]uuid.UUID, error)
 	UpdateChatLabelsByID(ctx context.Context, arg UpdateChatLabelsByIDParams) (Chat, error)
 	// Updates the cached injected context parts (AGENTS.md +
 	// skills) on the chat row. Called only when context changes
@@ -927,6 +945,7 @@ type sqlcQuerier interface {
 	UpdateUsageEventsPostPublish(ctx context.Context, arg UpdateUsageEventsPostPublishParams) error
 	UpdateUserChatCompactionThreshold(ctx context.Context, arg UpdateUserChatCompactionThresholdParams) (UserConfig, error)
 	UpdateUserChatCustomPrompt(ctx context.Context, arg UpdateUserChatCustomPromptParams) (UserConfig, error)
+	UpdateUserChatProviderKey(ctx context.Context, arg UpdateUserChatProviderKeyParams) (UserChatProviderKey, error)
 	UpdateUserDeletedByID(ctx context.Context, id uuid.UUID) error
 	UpdateUserGithubComUserID(ctx context.Context, arg UpdateUserGithubComUserIDParams) error
 	UpdateUserHashedOneTimePasscode(ctx context.Context, arg UpdateUserHashedOneTimePasscodeParams) error
@@ -988,7 +1007,6 @@ type sqlcQuerier interface {
 	UpsertChatUsageLimitGroupOverride(ctx context.Context, arg UpsertChatUsageLimitGroupOverrideParams) (UpsertChatUsageLimitGroupOverrideRow, error)
 	UpsertChatUsageLimitUserOverride(ctx context.Context, arg UpsertChatUsageLimitUserOverrideParams) (UpsertChatUsageLimitUserOverrideRow, error)
 	UpsertChatWorkspaceTTL(ctx context.Context, workspaceTtl string) error
-	UpsertConnectionLog(ctx context.Context, arg UpsertConnectionLogParams) (ConnectionLog, error)
 	// The default proxy is implied and not actually stored in the database.
 	// So we need to store it's configuration here for display purposes.
 	// The functional values are immutable and controlled implicitly.
@@ -1015,6 +1033,7 @@ type sqlcQuerier interface {
 	// used to store the data, and the minutes are summed for each user and template
 	// combination. The result is stored in the template_usage_stats table.
 	UpsertTemplateUsageStats(ctx context.Context) error
+	UpsertUserChatProviderKey(ctx context.Context, arg UpsertUserChatProviderKeyParams) (UserChatProviderKey, error)
 	UpsertWebpushVAPIDKeys(ctx context.Context, arg UpsertWebpushVAPIDKeysParams) error
 	UpsertWorkspaceAgentPortShare(ctx context.Context, arg UpsertWorkspaceAgentPortShareParams) (WorkspaceAgentPortShare, error)
 	UpsertWorkspaceApp(ctx context.Context, arg UpsertWorkspaceAppParams) (WorkspaceApp, error)

coderd/database/querier_test.go

@@ -1261,10 +1261,11 @@ func TestGetAuthorizedChats(t *testing.T) {
 	// Create FK dependencies: a chat provider and model config.
 	ctx := testutil.Context(t, testutil.WaitMedium)
 	_, err = db.InsertChatProvider(ctx, database.InsertChatProviderParams{
-		Provider:    "openai",
-		DisplayName: "OpenAI",
-		APIKey:      "test-key",
-		Enabled:     true,
+		Provider:             "openai",
+		DisplayName:          "OpenAI",
+		APIKey:               "test-key",
+		Enabled:              true,
+		CentralApiKeyEnabled: true,
 	})
 	require.NoError(t, err)
 
@@ -1285,6 +1286,7 @@ func TestGetAuthorizedChats(t *testing.T) {
 	// Create 3 chats owned by owner.
 	for i := range 3 {
 		_, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           owner.ID,
 			LastModelConfigID: modelCfg.ID,
 			Title:             fmt.Sprintf("owner chat %d", i+1),
@@ -1295,6 +1297,7 @@ func TestGetAuthorizedChats(t *testing.T) {
 	// Create 2 chats owned by member.
 	for i := range 2 {
 		_, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           member.ID,
 			LastModelConfigID: modelCfg.ID,
 			Title:             fmt.Sprintf("member chat %d", i+1),
@@ -1416,6 +1419,7 @@ func TestGetAuthorizedChats(t *testing.T) {
 		})
 		for i := range 7 {
 			_, err := db.InsertChat(ctx, database.InsertChatParams{
+				Status:            database.ChatStatusWaiting,
 				OwnerID:           paginationUser.ID,
 				LastModelConfigID: modelCfg.ID,
 				Title:             fmt.Sprintf("pagination chat %d", i+1),
@@ -3562,9 +3566,11 @@ func connectionOnlyIDs[T database.ConnectionLog | database.GetConnectionLogsOffs
 	return ids
 }
 
-func TestUpsertConnectionLog(t *testing.T) {
+func TestBatchUpsertConnectionLogs(t *testing.T) {
 	t.Parallel()
+
 	createWorkspace := func(t *testing.T, db database.Store) database.WorkspaceTable {
+		t.Helper()
 		u := dbgen.User(t, db, database.User{})
 		o := dbgen.Organization(t, db, database.Organization{})
 		tpl := dbgen.Template(t, db, database.Template{
@@ -3580,253 +3586,536 @@ func TestUpsertConnectionLog(t *testing.T) {
 		})
 	}
 
+	// zeroTime is the sentinel value that the SQL treats as "no
+	// connect/disconnect time provided".
+	zeroTime := time.Time{}
+
+	defaultIP := pqtype.Inet{
+		IPNet: net.IPNet{
+			IP:   net.IPv4(127, 0, 0, 1),
+			Mask: net.IPv4Mask(255, 255, 255, 255),
+		},
+		Valid: true,
+	}
+
+	t.Run("SingleConnect", func(t *testing.T) {
+		t.Parallel()
+		db, _ := dbtestutil.NewDB(t)
+		ctx := context.Background()
+		ws := createWorkspace(t, db)
+		connID := uuid.New()
+		connectTime := dbtime.Now()
+
+		err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{connectTime},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{0},
+			CodeValid:        []bool{false},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{""},
+			DisconnectTime:   []time.Time{zeroTime},
+		})
+		require.NoError(t, err)
+
+		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows, 1)
+		require.True(t, connectTime.Equal(rows[0].ConnectionLog.ConnectTime))
+		require.False(t, rows[0].ConnectionLog.DisconnectTime.Valid,
+			"disconnect_time should be NULL for a connect-only event")
+	})
+
 	t.Run("ConnectThenDisconnect", func(t *testing.T) {
 		t.Parallel()
 		db, _ := dbtestutil.NewDB(t)
 		ctx := context.Background()
-
 		ws := createWorkspace(t, db)
-
-		connectionID := uuid.New()
-		agentName := "test-agent"
-
-		// 1. Insert a 'connect' event.
+		connID := uuid.New()
 		connectTime := dbtime.Now()
-		connectParams := database.UpsertConnectionLogParams{
-			ID:               uuid.New(),
-			Time:             connectTime,
-			OrganizationID:   ws.OrganizationID,
-			WorkspaceOwnerID: ws.OwnerID,
-			WorkspaceID:      ws.ID,
-			WorkspaceName:    ws.Name,
-			AgentName:        agentName,
-			Type:             database.ConnectionTypeSsh,
-			ConnectionID:     uuid.NullUUID{UUID: connectionID, Valid: true},
-			ConnectionStatus: database.ConnectionStatusConnected,
-			Ip: pqtype.Inet{
-				IPNet: net.IPNet{
-					IP:   net.IPv4(127, 0, 0, 1),
-					Mask: net.IPv4Mask(255, 255, 255, 255),
-				},
-				Valid: true,
-			},
-		}
 
-		log1, err := db.UpsertConnectionLog(ctx, connectParams)
+		// Insert connect.
+		err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{connectTime},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{0},
+			CodeValid:        []bool{false},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{""},
+			DisconnectTime:   []time.Time{zeroTime},
+		})
+		require.NoError(t, err)
+
+		// Insert disconnect for same connection.
+		disconnectTime := connectTime.Add(time.Second)
+		err = db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{zeroTime},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{1},
+			CodeValid:        []bool{true},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{"test disconnect"},
+			DisconnectTime:   []time.Time{disconnectTime},
+		})
 		require.NoError(t, err)
-		require.Equal(t, connectParams.ID, log1.ID)
-		require.False(t, log1.DisconnectTime.Valid, "DisconnectTime should not be set on connect")
 
-		// Check that one row exists.
 		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
 		require.NoError(t, err)
 		require.Len(t, rows, 1)
-
-		// 2. Insert a 'disconnected' event for the same connection.
-		disconnectTime := connectTime.Add(time.Second)
-		disconnectParams := database.UpsertConnectionLogParams{
-			ConnectionID:     uuid.NullUUID{UUID: connectionID, Valid: true},
-			WorkspaceID:      ws.ID,
-			AgentName:        agentName,
-			ConnectionStatus: database.ConnectionStatusDisconnected,
-
-			// Updated to:
-			Time:             disconnectTime,
-			DisconnectReason: sql.NullString{String: "test disconnect", Valid: true},
-			Code:             sql.NullInt32{Int32: 1, Valid: true},
-
-			// Ignored
-			ID:               uuid.New(),
-			OrganizationID:   ws.OrganizationID,
-			WorkspaceOwnerID: ws.OwnerID,
-			WorkspaceName:    ws.Name,
-			Type:             database.ConnectionTypeSsh,
-			Ip: pqtype.Inet{
-				IPNet: net.IPNet{
-					IP:   net.IPv4(127, 0, 0, 1),
-					Mask: net.IPv4Mask(255, 255, 255, 254),
-				},
-				Valid: true,
-			},
-		}
-
-		log2, err := db.UpsertConnectionLog(ctx, disconnectParams)
-		require.NoError(t, err)
-
-		// Updated
-		require.Equal(t, log1.ID, log2.ID)
-		require.True(t, log2.DisconnectTime.Valid)
-		require.True(t, disconnectTime.Equal(log2.DisconnectTime.Time))
-		require.Equal(t, disconnectParams.DisconnectReason.String, log2.DisconnectReason.String)
-
-		rows, err = db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{})
-		require.NoError(t, err)
-		require.Len(t, rows, 1)
+		row := rows[0].ConnectionLog
+		require.True(t, connectTime.Equal(row.ConnectTime))
+		require.True(t, row.DisconnectTime.Valid)
+		require.True(t, disconnectTime.Equal(row.DisconnectTime.Time))
+		require.Equal(t, "test disconnect", row.DisconnectReason.String)
+		require.Equal(t, int32(1), row.Code.Int32)
 	})
 
-	t.Run("ConnectDoesNotUpdate", func(t *testing.T) {
+	t.Run("DuplicateConnectIsNoOp", func(t *testing.T) {
 		t.Parallel()
 		db, _ := dbtestutil.NewDB(t)
 		ctx := context.Background()
-
 		ws := createWorkspace(t, db)
-
-		connectionID := uuid.New()
-		agentName := "test-agent"
-
-		// 1. Insert a 'connect' event.
+		connID := uuid.New()
 		connectTime := dbtime.Now()
-		connectParams := database.UpsertConnectionLogParams{
-			ID:               uuid.New(),
-			Time:             connectTime,
-			OrganizationID:   ws.OrganizationID,
-			WorkspaceOwnerID: ws.OwnerID,
-			WorkspaceID:      ws.ID,
-			WorkspaceName:    ws.Name,
-			AgentName:        agentName,
-			Type:             database.ConnectionTypeSsh,
-			ConnectionID:     uuid.NullUUID{UUID: connectionID, Valid: true},
-			ConnectionStatus: database.ConnectionStatusConnected,
-			Ip: pqtype.Inet{
-				IPNet: net.IPNet{
-					IP:   net.IPv4(127, 0, 0, 1),
-					Mask: net.IPv4Mask(255, 255, 255, 255),
-				},
-				Valid: true,
-			},
+
+		mkParams := func(ct time.Time, ip pqtype.Inet) database.BatchUpsertConnectionLogsParams {
+			return database.BatchUpsertConnectionLogsParams{
+				ID:               []uuid.UUID{uuid.New()},
+				ConnectTime:      []time.Time{ct},
+				OrganizationID:   []uuid.UUID{ws.OrganizationID},
+				WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+				WorkspaceID:      []uuid.UUID{ws.ID},
+				WorkspaceName:    []string{ws.Name},
+				AgentName:        []string{"agent"},
+				Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+				Code:             []int32{0},
+				CodeValid:        []bool{false},
+				Ip:               []pqtype.Inet{ip},
+				UserAgent:        []string{""},
+				UserID:           []uuid.UUID{uuid.Nil},
+				SlugOrPort:       []string{""},
+				ConnectionID:     []uuid.UUID{connID},
+				DisconnectReason: []string{""},
+				DisconnectTime:   []time.Time{zeroTime},
+			}
 		}
 
-		log, err := db.UpsertConnectionLog(ctx, connectParams)
+		err := db.BatchUpsertConnectionLogs(ctx, mkParams(connectTime, defaultIP))
 		require.NoError(t, err)
 
-		// 2. Insert another 'connect' event for the same connection.
-		connectTime2 := connectTime.Add(time.Second)
-		connectParams2 := database.UpsertConnectionLogParams{
-			ConnectionID:     uuid.NullUUID{UUID: connectionID, Valid: true},
-			WorkspaceID:      ws.ID,
-			AgentName:        agentName,
-			ConnectionStatus: database.ConnectionStatusConnected,
+		rows1, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows1, 1)
 
-			// Ignored
-			ID:               uuid.New(),
-			Time:             connectTime2,
-			OrganizationID:   ws.OrganizationID,
-			WorkspaceOwnerID: ws.OwnerID,
-			WorkspaceName:    ws.Name,
-			Type:             database.ConnectionTypeSsh,
-			Code:             sql.NullInt32{Int32: 0, Valid: false},
-			Ip: pqtype.Inet{
-				IPNet: net.IPNet{
-					IP:   net.IPv4(127, 0, 0, 1),
-					Mask: net.IPv4Mask(255, 255, 255, 254),
-				},
-				Valid: true,
+		// Second connect with later time and different IP.
+		otherIP := pqtype.Inet{
+			IPNet: net.IPNet{
+				IP:   net.IPv4(10, 0, 0, 1),
+				Mask: net.IPv4Mask(255, 255, 255, 255),
 			},
+			Valid: true,
 		}
-
-		origLog, err := db.UpsertConnectionLog(ctx, connectParams2)
+		err = db.BatchUpsertConnectionLogs(ctx, mkParams(connectTime.Add(time.Second), otherIP))
 		require.NoError(t, err)
-		require.Equal(t, log, origLog, "connect update should be a no-op")
 
-		// Check that still only one row exists.
-		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{})
+		rows2, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
 		require.NoError(t, err)
-		require.Len(t, rows, 1)
-		require.Equal(t, log, rows[0].ConnectionLog)
+		require.Len(t, rows2, 1)
+
+		// The LEAST logic should pick the earlier connect_time; IP and
+		// other fields are not updated on conflict.
+		require.True(t, connectTime.Equal(rows2[0].ConnectionLog.ConnectTime),
+			"connect_time should remain the original (earlier) value")
 	})
 
-	t.Run("DisconnectThenConnect", func(t *testing.T) {
+	t.Run("OrderIndependentConnectTime", func(t *testing.T) {
 		t.Parallel()
-
 		db, _ := dbtestutil.NewDB(t)
 		ctx := context.Background()
-
 		ws := createWorkspace(t, db)
-
-		connectionID := uuid.New()
-		agentName := "test-agent"
-
-		// Insert just a 'disconect' event
+		connID := uuid.New()
 		disconnectTime := dbtime.Now()
-		disconnectParams := database.UpsertConnectionLogParams{
-			ID:               uuid.New(),
-			Time:             disconnectTime,
-			OrganizationID:   ws.OrganizationID,
-			WorkspaceOwnerID: ws.OwnerID,
-			WorkspaceID:      ws.ID,
-			WorkspaceName:    ws.Name,
-			AgentName:        agentName,
-			Type:             database.ConnectionTypeSsh,
-			ConnectionID:     uuid.NullUUID{UUID: connectionID, Valid: true},
-			ConnectionStatus: database.ConnectionStatusDisconnected,
-			DisconnectReason: sql.NullString{String: "server shutting down", Valid: true},
-			Ip: pqtype.Inet{
-				IPNet: net.IPNet{
-					IP:   net.IPv4(127, 0, 0, 1),
-					Mask: net.IPv4Mask(255, 255, 255, 255),
-				},
-				Valid: true,
-			},
+		connectTime := disconnectTime.Add(-5 * time.Second)
+
+		// Disconnect arrives first.
+		err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{disconnectTime},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{0},
+			CodeValid:        []bool{true},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{"bye"},
+			DisconnectTime:   []time.Time{disconnectTime},
+		})
+		require.NoError(t, err)
+
+		// Connect arrives second with the real (earlier) connect_time.
+		err = db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{connectTime},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{0},
+			CodeValid:        []bool{false},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{""},
+			DisconnectTime:   []time.Time{zeroTime},
+		})
+		require.NoError(t, err)
+
+		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows, 1)
+		require.True(t, connectTime.Equal(rows[0].ConnectionLog.ConnectTime),
+			"LEAST should pick the earlier connect_time")
+	})
+
+	t.Run("DisconnectFieldsAreWriteOnce", func(t *testing.T) {
+		t.Parallel()
+		db, _ := dbtestutil.NewDB(t)
+		ctx := context.Background()
+		ws := createWorkspace(t, db)
+		connID := uuid.New()
+		disconnectTime := dbtime.Now()
+
+		mkDisconnect := func(reason string, code int32) database.BatchUpsertConnectionLogsParams {
+			return database.BatchUpsertConnectionLogsParams{
+				ID:               []uuid.UUID{uuid.New()},
+				ConnectTime:      []time.Time{disconnectTime},
+				OrganizationID:   []uuid.UUID{ws.OrganizationID},
+				WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+				WorkspaceID:      []uuid.UUID{ws.ID},
+				WorkspaceName:    []string{ws.Name},
+				AgentName:        []string{"agent"},
+				Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+				Code:             []int32{code},
+				CodeValid:        []bool{true},
+				Ip:               []pqtype.Inet{defaultIP},
+				UserAgent:        []string{""},
+				UserID:           []uuid.UUID{uuid.Nil},
+				SlugOrPort:       []string{""},
+				ConnectionID:     []uuid.UUID{connID},
+				DisconnectReason: []string{reason},
+				DisconnectTime:   []time.Time{disconnectTime},
+			}
 		}
 
-		_, err := db.UpsertConnectionLog(ctx, disconnectParams)
+		err := db.BatchUpsertConnectionLogs(ctx, mkDisconnect("first reason", 1))
 		require.NoError(t, err)
 
-		firstRows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{})
+		// Second disconnect with different reason and code.
+		err = db.BatchUpsertConnectionLogs(ctx, mkDisconnect("second reason", 2))
 		require.NoError(t, err)
-		require.Len(t, firstRows, 1)
 
-		// We expect the connection event to be marked as closed with the start
-		// and close time being the same.
-		require.True(t, firstRows[0].ConnectionLog.DisconnectTime.Valid)
-		require.Equal(t, disconnectTime, firstRows[0].ConnectionLog.DisconnectTime.Time.UTC())
-		require.Equal(t, firstRows[0].ConnectionLog.ConnectTime.UTC(), firstRows[0].ConnectionLog.DisconnectTime.Time.UTC())
+		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows, 1)
+		row := rows[0].ConnectionLog
+		require.Equal(t, "first reason", row.DisconnectReason.String,
+			"disconnect_reason should not be overwritten")
+		require.Equal(t, int32(1), row.Code.Int32,
+			"code should not be overwritten")
+	})
 
-		// Now insert a 'connect' event for the same connection.
-		// This should be a no op
-		connectTime := disconnectTime.Add(time.Second)
-		connectParams := database.UpsertConnectionLogParams{
-			ID:               uuid.New(),
-			Time:             connectTime,
-			OrganizationID:   ws.OrganizationID,
-			WorkspaceOwnerID: ws.OwnerID,
-			WorkspaceID:      ws.ID,
-			WorkspaceName:    ws.Name,
-			AgentName:        agentName,
-			Type:             database.ConnectionTypeSsh,
-			ConnectionID:     uuid.NullUUID{UUID: connectionID, Valid: true},
-			ConnectionStatus: database.ConnectionStatusConnected,
-			DisconnectReason: sql.NullString{String: "reconnected", Valid: true},
-			Code:             sql.NullInt32{Int32: 0, Valid: false},
-			Ip: pqtype.Inet{
-				IPNet: net.IPNet{
-					IP:   net.IPv4(127, 0, 0, 1),
-					Mask: net.IPv4Mask(255, 255, 255, 255),
-				},
-				Valid: true,
-			},
+	t.Run("ConnectAfterDisconnectIsNoOp", func(t *testing.T) {
+		t.Parallel()
+		db, _ := dbtestutil.NewDB(t)
+		ctx := context.Background()
+		ws := createWorkspace(t, db)
+		connID := uuid.New()
+		disconnectTime := dbtime.Now()
+
+		// Insert disconnect first.
+		err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{disconnectTime},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{42},
+			CodeValid:        []bool{true},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{"server shutdown"},
+			DisconnectTime:   []time.Time{disconnectTime},
+		})
+		require.NoError(t, err)
+
+		rows1, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows1, 1)
+		require.True(t, rows1[0].ConnectionLog.DisconnectTime.Valid)
+		require.Equal(t, "server shutdown", rows1[0].ConnectionLog.DisconnectReason.String)
+		require.Equal(t, int32(42), rows1[0].ConnectionLog.Code.Int32)
+
+		// Insert connect for same connection_id.
+		err = db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{disconnectTime.Add(time.Second)},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{0},
+			CodeValid:        []bool{false},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{""},
+			DisconnectTime:   []time.Time{zeroTime},
+		})
+		require.NoError(t, err)
+
+		rows2, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows2, 1)
+		row := rows2[0].ConnectionLog
+		require.True(t, row.DisconnectTime.Valid,
+			"disconnect_time should not be cleared by a later connect")
+		require.Equal(t, "server shutdown", row.DisconnectReason.String,
+			"disconnect_reason should not be cleared")
+		require.Equal(t, int32(42), row.Code.Int32,
+			"code should not be cleared")
+	})
+
+	t.Run("CodeZeroPreserved", func(t *testing.T) {
+		t.Parallel()
+		db, _ := dbtestutil.NewDB(t)
+		ctx := context.Background()
+		ws := createWorkspace(t, db)
+		connID := uuid.New()
+		now := dbtime.Now()
+
+		err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{now},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{0},
+			CodeValid:        []bool{true},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{"normal"},
+			DisconnectTime:   []time.Time{now},
+		})
+		require.NoError(t, err)
+
+		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows, 1)
+		require.True(t, rows[0].ConnectionLog.Code.Valid, "code should be non-NULL")
+		require.Equal(t, int32(0), rows[0].ConnectionLog.Code.Int32,
+			"code=0 should be preserved, not treated as NULL")
+	})
+
+	t.Run("CodeNullWhenInvalid", func(t *testing.T) {
+		t.Parallel()
+		db, _ := dbtestutil.NewDB(t)
+		ctx := context.Background()
+		ws := createWorkspace(t, db)
+		connID := uuid.New()
+		now := dbtime.Now()
+
+		err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               []uuid.UUID{uuid.New()},
+			ConnectTime:      []time.Time{now},
+			OrganizationID:   []uuid.UUID{ws.OrganizationID},
+			WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+			WorkspaceID:      []uuid.UUID{ws.ID},
+			WorkspaceName:    []string{ws.Name},
+			AgentName:        []string{"agent"},
+			Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+			Code:             []int32{99},
+			CodeValid:        []bool{false},
+			Ip:               []pqtype.Inet{defaultIP},
+			UserAgent:        []string{""},
+			UserID:           []uuid.UUID{uuid.Nil},
+			SlugOrPort:       []string{""},
+			ConnectionID:     []uuid.UUID{connID},
+			DisconnectReason: []string{""},
+			DisconnectTime:   []time.Time{zeroTime},
+		})
+		require.NoError(t, err)
+
+		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
+		require.NoError(t, err)
+		require.Len(t, rows, 1)
+		require.False(t, rows[0].ConnectionLog.Code.Valid,
+			"code should be NULL when code_valid is false")
+	})
+
+	t.Run("NullConnectionIDEvents", func(t *testing.T) {
+		t.Parallel()
+		db, _ := dbtestutil.NewDB(t)
+		ctx := context.Background()
+		ws := createWorkspace(t, db)
+		now := dbtime.Now()
+
+		// Insert two web events with NULL connection_id (uuid.Nil →
+		// NULL via NULLIF) for the same workspace/agent.
+		for i := range 2 {
+			err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+				ID:               []uuid.UUID{uuid.New()},
+				ConnectTime:      []time.Time{now.Add(time.Duration(i) * time.Second)},
+				OrganizationID:   []uuid.UUID{ws.OrganizationID},
+				WorkspaceOwnerID: []uuid.UUID{ws.OwnerID},
+				WorkspaceID:      []uuid.UUID{ws.ID},
+				WorkspaceName:    []string{ws.Name},
+				AgentName:        []string{"agent"},
+				Type:             []database.ConnectionType{database.ConnectionTypeSsh},
+				Code:             []int32{200},
+				CodeValid:        []bool{true},
+				Ip:               []pqtype.Inet{defaultIP},
+				UserAgent:        []string{"Mozilla/5.0"},
+				UserID:           []uuid.UUID{uuid.Nil},
+				SlugOrPort:       []string{"web-terminal"},
+				ConnectionID:     []uuid.UUID{uuid.Nil},
+				DisconnectReason: []string{""},
+				DisconnectTime:   []time.Time{zeroTime},
+			})
+			require.NoError(t, err)
 		}
 
-		_, err = db.UpsertConnectionLog(ctx, connectParams)
+		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
 		require.NoError(t, err)
+		require.Len(t, rows, 2,
+			"NULL connection_id rows should not conflict with each other")
+	})
 
-		secondRows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{})
-		require.NoError(t, err)
-		require.Len(t, secondRows, 1)
-		require.Equal(t, firstRows, secondRows)
+	t.Run("MultipleIndependentConnections", func(t *testing.T) {
+		t.Parallel()
+		db, _ := dbtestutil.NewDB(t)
+		ctx := context.Background()
+		ws := createWorkspace(t, db)
+		now := dbtime.Now()
 
-		// Upsert a disconnection, which should also be a no op
-		disconnectParams.DisconnectReason = sql.NullString{
-			String: "updated close reason",
-			Valid:  true,
+		n := 5
+		ids := make([]uuid.UUID, n)
+		connectTimes := make([]time.Time, n)
+		orgIDs := make([]uuid.UUID, n)
+		ownerIDs := make([]uuid.UUID, n)
+		wsIDs := make([]uuid.UUID, n)
+		wsNames := make([]string, n)
+		agentNames := make([]string, n)
+		types := make([]database.ConnectionType, n)
+		codes := make([]int32, n)
+		codeValids := make([]bool, n)
+		ips := make([]pqtype.Inet, n)
+		userAgents := make([]string, n)
+		userIDs := make([]uuid.UUID, n)
+		slugOrPorts := make([]string, n)
+		connIDs := make([]uuid.UUID, n)
+		disconnectReasons := make([]string, n)
+		disconnectTimes := make([]time.Time, n)
+
+		for i := range n {
+			ids[i] = uuid.New()
+			connectTimes[i] = now.Add(time.Duration(i) * time.Second)
+			orgIDs[i] = ws.OrganizationID
+			ownerIDs[i] = ws.OwnerID
+			wsIDs[i] = ws.ID
+			wsNames[i] = ws.Name
+			agentNames[i] = "agent"
+			types[i] = database.ConnectionTypeSsh
+			codes[i] = 0
+			codeValids[i] = false
+			ips[i] = defaultIP
+			userAgents[i] = ""
+			userIDs[i] = uuid.Nil
+			slugOrPorts[i] = ""
+			connIDs[i] = uuid.New()
+			disconnectReasons[i] = ""
+			disconnectTimes[i] = zeroTime
 		}
-		_, err = db.UpsertConnectionLog(ctx, disconnectParams)
+
+		err := db.BatchUpsertConnectionLogs(ctx, database.BatchUpsertConnectionLogsParams{
+			ID:               ids,
+			ConnectTime:      connectTimes,
+			OrganizationID:   orgIDs,
+			WorkspaceOwnerID: ownerIDs,
+			WorkspaceID:      wsIDs,
+			WorkspaceName:    wsNames,
+			AgentName:        agentNames,
+			Type:             types,
+			Code:             codes,
+			CodeValid:        codeValids,
+			Ip:               ips,
+			UserAgent:        userAgents,
+			UserID:           userIDs,
+			SlugOrPort:       slugOrPorts,
+			ConnectionID:     connIDs,
+			DisconnectReason: disconnectReasons,
+			DisconnectTime:   disconnectTimes,
+		})
 		require.NoError(t, err)
-		thirdRows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{})
+
+		rows, err := db.GetConnectionLogsOffset(ctx, database.GetConnectionLogsOffsetParams{LimitOpt: 10})
 		require.NoError(t, err)
-		require.Len(t, secondRows, 1)
-		// The close reason shouldn't be updated
-		require.Equal(t, secondRows, thirdRows)
+		require.Len(t, rows, n, "each unique connection_id should produce its own row")
 	})
 }
 
@@ -9453,10 +9742,11 @@ func TestInsertChatMessages(t *testing.T) {
 		provider := "openai"
 
 		_, err := store.InsertChatProvider(ctx, database.InsertChatProviderParams{
-			Provider:    provider,
-			DisplayName: "OpenAI",
-			APIKey:      "test-key",
-			Enabled:     true,
+			Provider:             provider,
+			DisplayName:          "OpenAI",
+			APIKey:               "test-key",
+			Enabled:              true,
+			CentralApiKeyEnabled: true,
 		})
 		require.NoError(t, err)
 
@@ -9472,6 +9762,7 @@ func TestInsertChatMessages(t *testing.T) {
 		)
 
 		chat, err := store.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           user.ID,
 			LastModelConfigID: modelConfigA.ID,
 			Title:             "test-chat-" + uuid.NewString(),
@@ -9617,10 +9908,11 @@ func TestGetChatMessagesForPromptByChatID(t *testing.T) {
 
 	// A chat_providers row is required as a FK for model configs.
 	_, err := db.InsertChatProvider(ctx, database.InsertChatProviderParams{
-		Provider:    "openai",
-		DisplayName: "OpenAI",
-		APIKey:      "test-key",
-		Enabled:     true,
+		Provider:             "openai",
+		DisplayName:          "OpenAI",
+		APIKey:               "test-key",
+		Enabled:              true,
+		CentralApiKeyEnabled: true,
 	})
 	require.NoError(t, err)
 
@@ -9641,6 +9933,7 @@ func TestGetChatMessagesForPromptByChatID(t *testing.T) {
 	newChat := func(t *testing.T) database.Chat {
 		t.Helper()
 		chat, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           user.ID,
 			LastModelConfigID: modelCfg.ID,
 			Title:             "test-chat-" + uuid.NewString(),
@@ -9987,10 +10280,11 @@ func TestGetPRInsights(t *testing.T) {
 		user := dbgen.User(t, store, database.User{})
 
 		_, err := store.InsertChatProvider(ctx, database.InsertChatProviderParams{
-			Provider:    "anthropic",
-			DisplayName: "Anthropic",
-			APIKey:      "test-key",
-			Enabled:     true,
+			Provider:             "anthropic",
+			DisplayName:          "Anthropic",
+			APIKey:               "test-key",
+			Enabled:              true,
+			CentralApiKeyEnabled: true,
 		})
 		require.NoError(t, err)
 
@@ -10014,6 +10308,7 @@ func TestGetPRInsights(t *testing.T) {
 	createChat := func(t *testing.T, store database.Store, userID, mcID uuid.UUID, title string) database.Chat {
 		t.Helper()
 		chat, err := store.InsertChat(context.Background(), database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           userID,
 			LastModelConfigID: mcID,
 			Title:             title,
@@ -10149,6 +10444,7 @@ func TestGetPRInsights(t *testing.T) {
 	createChildChat := func(t *testing.T, store database.Store, userID, mcID, parentID, rootID uuid.UUID, title string) database.Chat {
 		t.Helper()
 		chat, err := store.InsertChat(context.Background(), database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           userID,
 			LastModelConfigID: mcID,
 			Title:             title,
@@ -10509,10 +10805,11 @@ func TestChatPinOrderQueries(t *testing.T) {
 		// timed test context doesn't tick during DB init.
 		bg := context.Background()
 		_, err := db.InsertChatProvider(bg, database.InsertChatProviderParams{
-			Provider:    "openai",
-			DisplayName: "OpenAI",
-			APIKey:      "test-key",
-			Enabled:     true,
+			Provider:             "openai",
+			DisplayName:          "OpenAI",
+			APIKey:               "test-key",
+			Enabled:              true,
+			CentralApiKeyEnabled: true,
 		})
 		require.NoError(t, err)
 
@@ -10538,6 +10835,7 @@ func TestChatPinOrderQueries(t *testing.T) {
 		t.Helper()
 
 		chat, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           ownerID,
 			LastModelConfigID: modelCfgID,
 			Title:             title,
@@ -10646,7 +10944,8 @@ func TestChatPinOrderQueries(t *testing.T) {
 		}
 
 		// Archive the middle pin.
-		require.NoError(t, db.ArchiveChatByID(ctx, second.ID))
+		_, err := db.ArchiveChatByID(ctx, second.ID)
+		require.NoError(t, err)
 
 		// Archived chat should have pin_order cleared. Remaining
 		// pins keep their original positions; the next mutation
@@ -10687,10 +10986,11 @@ func TestChatLabels(t *testing.T) {
 	owner := dbgen.User(t, db, database.User{})
 
 	_, err = db.InsertChatProvider(ctx, database.InsertChatProviderParams{
-		Provider:    "openai",
-		DisplayName: "OpenAI",
-		APIKey:      "test-key",
-		Enabled:     true,
+		Provider:             "openai",
+		DisplayName:          "OpenAI",
+		APIKey:               "test-key",
+		Enabled:              true,
+		CentralApiKeyEnabled: true,
 	})
 	require.NoError(t, err)
 
@@ -10717,6 +11017,7 @@ func TestChatLabels(t *testing.T) {
 		require.NoError(t, err)
 
 		chat, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           owner.ID,
 			LastModelConfigID: modelCfg.ID,
 			Title:             "labeled-chat",
@@ -10739,6 +11040,7 @@ func TestChatLabels(t *testing.T) {
 		ctx := testutil.Context(t, testutil.WaitMedium)
 
 		chat, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           owner.ID,
 			LastModelConfigID: modelCfg.ID,
 			Title:             "no-labels-chat",
@@ -10754,6 +11056,7 @@ func TestChatLabels(t *testing.T) {
 		ctx := testutil.Context(t, testutil.WaitMedium)
 
 		chat, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           owner.ID,
 			LastModelConfigID: modelCfg.ID,
 			Title:             "update-labels-chat",
@@ -10794,6 +11097,7 @@ func TestChatLabels(t *testing.T) {
 		require.NoError(t, err)
 
 		chat, err := db.InsertChat(ctx, database.InsertChatParams{
+			Status:            database.ChatStatusWaiting,
 			OwnerID:           owner.ID,
 			LastModelConfigID: modelCfg.ID,
 			Title:             "original-title",
@@ -10830,6 +11134,7 @@ func TestChatLabels(t *testing.T) {
 			labelsJSON, err := json.Marshal(tc.labels)
 			require.NoError(t, err)
 			_, err = db.InsertChat(ctx, database.InsertChatParams{
+				Status:            database.ChatStatusWaiting,
 				OwnerID:           owner.ID,
 				LastModelConfigID: modelCfg.ID,
 				Title:             tc.title,
@@ -10893,10 +11198,11 @@ func TestChatHasUnread(t *testing.T) {
 	user := dbgen.User(t, store, database.User{})
 
 	_, err := store.InsertChatProvider(ctx, database.InsertChatProviderParams{
-		Provider:    "openai",
-		DisplayName: "OpenAI",
-		APIKey:      "test-key",
-		Enabled:     true,
+		Provider:             "openai",
+		DisplayName:          "OpenAI",
+		APIKey:               "test-key",
+		Enabled:              true,
+		CentralApiKeyEnabled: true,
 	})
 	require.NoError(t, err)
 
@@ -10915,6 +11221,7 @@ func TestChatHasUnread(t *testing.T) {
 	require.NoError(t, err)
 
 	chat, err := store.InsertChat(ctx, database.InsertChatParams{
+		Status:            database.ChatStatusWaiting,
 		OwnerID:           user.ID,
 		LastModelConfigID: modelCfg.ID,
 		Title:             "test-chat-" + uuid.NewString(),

coderd/database/queries/aibridge.sql

@@ -1,8 +1,8 @@
 -- name: InsertAIBridgeInterception :one
 INSERT INTO aibridge_interceptions (
-	id, api_key_id, initiator_id, provider, model, metadata, started_at, client, client_session_id, thread_parent_id, thread_root_id
+	id, api_key_id, initiator_id, provider, provider_name, model, metadata, started_at, client, client_session_id, thread_parent_id, thread_root_id
 ) VALUES (
-	@id, @api_key_id, @initiator_id, @provider, @model, COALESCE(@metadata::jsonb, '{}'::jsonb), @started_at, @client, sqlc.narg('client_session_id'), sqlc.narg('thread_parent_interception_id')::uuid, sqlc.narg('thread_root_interception_id')::uuid
+	@id, @api_key_id, @initiator_id, @provider, @provider_name, @model, COALESCE(@metadata::jsonb, '{}'::jsonb), @started_at, @client, sqlc.narg('client_session_id'), sqlc.narg('thread_parent_interception_id')::uuid, sqlc.narg('thread_root_interception_id')::uuid
 )
 RETURNING *;
 
@@ -31,9 +31,9 @@ WHERE aibridge_interceptions.id = (
 
 -- name: InsertAIBridgeTokenUsage :one
 INSERT INTO aibridge_token_usages (
-  id, interception_id, provider_response_id, input_tokens, output_tokens, metadata, created_at
+  id, interception_id, provider_response_id, input_tokens, output_tokens, cache_read_input_tokens, cache_write_input_tokens, metadata, created_at
 ) VALUES (
-  @id, @interception_id, @provider_response_id, @input_tokens, @output_tokens, COALESCE(@metadata::jsonb, '{}'::jsonb), @created_at
+  @id, @interception_id, @provider_response_id, @input_tokens, @output_tokens, @cache_read_input_tokens, @cache_write_input_tokens, COALESCE(@metadata::jsonb, '{}'::jsonb), @created_at
 )
 RETURNING *;
 
@@ -299,21 +299,8 @@ token_aggregates AS (
     SELECT
         COALESCE(SUM(tu.input_tokens), 0) AS token_count_input,
         COALESCE(SUM(tu.output_tokens), 0) AS token_count_output,
-        -- Cached tokens are stored in metadata JSON, extract if available.
-        -- Read tokens may be stored in:
-        -- - cache_read_input (Anthropic)
-        -- - prompt_cached (OpenAI)
-        COALESCE(SUM(
-            COALESCE((tu.metadata->>'cache_read_input')::bigint, 0) +
-            COALESCE((tu.metadata->>'prompt_cached')::bigint, 0)
-        ), 0) AS token_count_cached_read,
-        -- Written tokens may be stored in:
-        -- - cache_creation_input (Anthropic)
-        -- Note that cache_ephemeral_5m_input and cache_ephemeral_1h_input on
-        -- Anthropic are included in the cache_creation_input field.
-        COALESCE(SUM(
-            COALESCE((tu.metadata->>'cache_creation_input')::bigint, 0)
-        ), 0) AS token_count_cached_written,
+        COALESCE(SUM(tu.cache_read_input_tokens), 0) AS token_count_cached_read,
+        COALESCE(SUM(tu.cache_write_input_tokens), 0) AS token_count_cached_written,
         COUNT(tu.id) AS token_usages_count
     FROM
         interceptions_in_range i
@@ -552,6 +539,8 @@ SELECT
 	sp.threads,
 	COALESCE(st.input_tokens, 0)::bigint AS input_tokens,
 	COALESCE(st.output_tokens, 0)::bigint AS output_tokens,
+	COALESCE(st.cache_read_input_tokens, 0)::bigint AS cache_read_input_tokens,
+	COALESCE(st.cache_write_input_tokens, 0)::bigint AS cache_write_input_tokens,
 	COALESCE(slp.prompt, '') AS last_prompt
 FROM
 	session_page sp
@@ -573,7 +562,9 @@ LEFT JOIN LATERAL (
 	-- Aggregate tokens only for this session's interceptions.
 	SELECT
 		COALESCE(SUM(tu.input_tokens), 0)::bigint AS input_tokens,
-		COALESCE(SUM(tu.output_tokens), 0)::bigint AS output_tokens
+		COALESCE(SUM(tu.output_tokens), 0)::bigint AS output_tokens,
+		COALESCE(SUM(tu.cache_read_input_tokens), 0)::bigint AS cache_read_input_tokens,
+		COALESCE(SUM(tu.cache_write_input_tokens), 0)::bigint AS cache_write_input_tokens
 	FROM aibridge_token_usages tu
 	WHERE tu.interception_id = ANY(sr.interception_ids)
 ) st ON true

coderd/database/queries/auditlogs.sql

@@ -149,94 +149,105 @@ VALUES (
 RETURNING *;
 
 -- name: CountAuditLogs :one
-SELECT COUNT(*)
-FROM audit_logs
-	LEFT JOIN users ON audit_logs.user_id = users.id
-	LEFT JOIN organizations ON audit_logs.organization_id = organizations.id
-	-- First join on workspaces to get the initial workspace create
-	-- to workspace build 1 id. This is because the first create is
-	-- is a different audit log than subsequent starts.
-	LEFT JOIN workspaces ON audit_logs.resource_type = 'workspace'
-	AND audit_logs.resource_id = workspaces.id
-	-- Get the reason from the build if the resource type
-	-- is a workspace_build
-	LEFT JOIN workspace_builds wb_build ON audit_logs.resource_type = 'workspace_build'
-	AND audit_logs.resource_id = wb_build.id
-	-- Get the reason from the build #1 if this is the first
-	-- workspace create.
-	LEFT JOIN workspace_builds wb_workspace ON audit_logs.resource_type = 'workspace'
-	AND audit_logs.action = 'create'
-	AND workspaces.id = wb_workspace.workspace_id
-	AND wb_workspace.build_number = 1
-WHERE
-	-- Filter resource_type
-	CASE
-		WHEN @resource_type::text != '' THEN resource_type = @resource_type::resource_type
-		ELSE true
-	END
-	-- Filter resource_id
-	AND CASE
-		WHEN @resource_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN resource_id = @resource_id
-		ELSE true
-	END
-	-- Filter organization_id
-	AND CASE
-		WHEN @organization_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN audit_logs.organization_id = @organization_id
-		ELSE true
-	END
-	-- Filter by resource_target
-	AND CASE
-		WHEN @resource_target::text != '' THEN resource_target = @resource_target
-		ELSE true
-	END
-	-- Filter action
-	AND CASE
-		WHEN @action::text != '' THEN action = @action::audit_action
-		ELSE true
-	END
-	-- Filter by user_id
-	AND CASE
-		WHEN @user_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN user_id = @user_id
-		ELSE true
-	END
-	-- Filter by username
-	AND CASE
-		WHEN @username::text != '' THEN user_id = (
-			SELECT id
-			FROM users
-			WHERE lower(username) = lower(@username)
-				AND deleted = false
-		)
-		ELSE true
-	END
-	-- Filter by user_email
-	AND CASE
-		WHEN @email::text != '' THEN users.email = @email
-		ELSE true
-	END
-	-- Filter by date_from
-	AND CASE
-		WHEN @date_from::timestamp with time zone != '0001-01-01 00:00:00Z' THEN "time" >= @date_from
-		ELSE true
-	END
-	-- Filter by date_to
-	AND CASE
-		WHEN @date_to::timestamp with time zone != '0001-01-01 00:00:00Z' THEN "time" <= @date_to
-		ELSE true
-	END
-	-- Filter by build_reason
-	AND CASE
-		WHEN @build_reason::text != '' THEN COALESCE(wb_build.reason::text, wb_workspace.reason::text) = @build_reason
-		ELSE true
-	END
-	-- Filter request_id
-	AND CASE
-		WHEN @request_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN audit_logs.request_id = @request_id
-		ELSE true
-	END
-	-- Authorize Filter clause will be injected below in CountAuthorizedAuditLogs
-	-- @authorize_filter
-;
+SELECT COUNT(*) FROM (
+	SELECT 1
+	FROM audit_logs
+		LEFT JOIN users ON audit_logs.user_id = users.id
+		LEFT JOIN organizations ON audit_logs.organization_id = organizations.id
+		-- First join on workspaces to get the initial workspace create
+		-- to workspace build 1 id. This is because the first create is
+		-- is a different audit log than subsequent starts.
+		LEFT JOIN workspaces ON audit_logs.resource_type = 'workspace'
+		AND audit_logs.resource_id = workspaces.id
+		-- Get the reason from the build if the resource type
+		-- is a workspace_build
+		LEFT JOIN workspace_builds wb_build ON audit_logs.resource_type = 'workspace_build'
+		AND audit_logs.resource_id = wb_build.id
+		-- Get the reason from the build #1 if this is the first
+		-- workspace create.
+		LEFT JOIN workspace_builds wb_workspace ON audit_logs.resource_type = 'workspace'
+		AND audit_logs.action = 'create'
+		AND workspaces.id = wb_workspace.workspace_id
+		AND wb_workspace.build_number = 1
+	WHERE
+		-- Filter resource_type
+		CASE
+			WHEN @resource_type::text != '' THEN resource_type = @resource_type::resource_type
+			ELSE true
+		END
+		-- Filter resource_id
+		AND CASE
+			WHEN @resource_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN resource_id = @resource_id
+			ELSE true
+		END
+		-- Filter organization_id
+		AND CASE
+			WHEN @organization_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN audit_logs.organization_id = @organization_id
+			ELSE true
+		END
+		-- Filter by resource_target
+		AND CASE
+			WHEN @resource_target::text != '' THEN resource_target = @resource_target
+			ELSE true
+		END
+		-- Filter action
+		AND CASE
+			WHEN @action::text != '' THEN action = @action::audit_action
+			ELSE true
+		END
+		-- Filter by user_id
+		AND CASE
+			WHEN @user_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN user_id = @user_id
+			ELSE true
+		END
+		-- Filter by username
+		AND CASE
+			WHEN @username::text != '' THEN user_id = (
+				SELECT id
+				FROM users
+				WHERE lower(username) = lower(@username)
+					AND deleted = false
+			)
+			ELSE true
+		END
+		-- Filter by user_email
+		AND CASE
+			WHEN @email::text != '' THEN users.email = @email
+			ELSE true
+		END
+		-- Filter by date_from
+		AND CASE
+			WHEN @date_from::timestamp with time zone != '0001-01-01 00:00:00Z' THEN "time" >= @date_from
+			ELSE true
+		END
+		-- Filter by date_to
+		AND CASE
+			WHEN @date_to::timestamp with time zone != '0001-01-01 00:00:00Z' THEN "time" <= @date_to
+			ELSE true
+		END
+		-- Filter by build_reason
+		AND CASE
+			WHEN @build_reason::text != '' THEN COALESCE(wb_build.reason::text, wb_workspace.reason::text) = @build_reason
+			ELSE true
+		END
+		-- Filter request_id
+		AND CASE
+			WHEN @request_id::uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN audit_logs.request_id = @request_id
+			ELSE true
+		END
+		-- Authorize Filter clause will be injected below in CountAuthorizedAuditLogs
+		-- @authorize_filter
+	-- Avoid a slow scan on a large table with joins. The caller
+	-- passes the count cap and we add 1 so the frontend can detect
+	-- capping and show "... of N+". A cap of 0 means no limit (NULLIF
+	-- -> NULL + 1 = NULL).
+	-- NOTE: Parameterizing this so that we can easily change from,
+	-- e.g., 2000 to 5000. However, use literal NULL (or no LIMIT)
+	-- here if disabling the capping on a large table permanently.
+	-- This way the PG planner can plan parallel execution for
+	-- potential large wins.
+	LIMIT NULLIF(@count_cap::int, 0) + 1
+) AS limited_count;
 
 -- name: DeleteOldAuditLogConnectionEvents :exec
 DELETE FROM audit_logs

coderd/database/queries/chatfiles.sql

@@ -8,3 +8,13 @@ SELECT * FROM chat_files WHERE id = @id::uuid;
 
 -- name: GetChatFilesByIDs :many
 SELECT * FROM chat_files WHERE id = ANY(@ids::uuid[]);
+
+-- name: GetChatFileMetadataByChatID :many
+-- GetChatFileMetadataByChatID returns lightweight file metadata for
+-- all files linked to a chat. The data column is excluded to avoid
+-- loading file content.
+SELECT cf.id, cf.owner_id, cf.organization_id, cf.name, cf.mimetype, cf.created_at
+FROM chat_files cf
+JOIN chat_file_links cfl ON cfl.file_id = cf.id
+WHERE cfl.chat_id = @chat_id::uuid
+ORDER BY cf.created_at ASC;

coderd/database/queries/chatproviders.sql

@@ -40,7 +40,10 @@ INSERT INTO chat_providers (
     base_url,
     api_key_key_id,
     created_by,
-    enabled
+    enabled,
+    central_api_key_enabled,
+    allow_user_api_key,
+    allow_central_api_key_fallback
 ) VALUES (
     @provider::text,
     @display_name::text,
@@ -48,7 +51,10 @@ INSERT INTO chat_providers (
     @base_url::text,
     sqlc.narg('api_key_key_id')::text,
     sqlc.narg('created_by')::uuid,
-    @enabled::boolean
+    @enabled::boolean,
+    @central_api_key_enabled::boolean,
+    @allow_user_api_key::boolean,
+    @allow_central_api_key_fallback::boolean
 )
 RETURNING
     *;
@@ -62,6 +68,9 @@ SET
     base_url = @base_url::text,
     api_key_key_id = sqlc.narg('api_key_key_id')::text,
     enabled = @enabled::boolean,
+    central_api_key_enabled = @central_api_key_enabled::boolean,
+    allow_user_api_key = @allow_user_api_key::boolean,
+    allow_central_api_key_fallback = @allow_central_api_key_fallback::boolean,
     updated_at = NOW()
 WHERE
     id = @id::uuid

coderd/database/queries/chats.sql

@@ -1,9 +1,24 @@
--- name: ArchiveChatByID :exec
-UPDATE chats SET archived = true, pin_order = 0, updated_at = NOW()
-WHERE id = @id OR root_chat_id = @id;
+-- name: ArchiveChatByID :many
+WITH chats AS (
+    UPDATE chats
+    SET archived = true, pin_order = 0, updated_at = NOW()
+    WHERE id = @id::uuid OR root_chat_id = @id::uuid
+    RETURNING *
+)
+SELECT *
+FROM chats
+ORDER BY (id = @id::uuid) DESC, created_at ASC, id ASC;
 
--- name: UnarchiveChatByID :exec
-UPDATE chats SET archived = false, updated_at = NOW() WHERE id = @id::uuid;
+-- name: UnarchiveChatByID :many
+WITH chats AS (
+    UPDATE chats
+    SET archived = false, updated_at = NOW()
+    WHERE id = @id::uuid OR root_chat_id = @id::uuid
+    RETURNING *
+)
+SELECT *
+FROM chats
+ORDER BY (id = @id::uuid) DESC, created_at ASC, id ASC;
 
 -- name: PinChatByID :exec
 WITH target_chat AS (
@@ -377,6 +392,7 @@ INSERT INTO chats (
     last_model_config_id,
     title,
     mode,
+    status,
     mcp_server_ids,
     labels
 ) VALUES (
@@ -389,6 +405,7 @@ INSERT INTO chats (
     @last_model_config_id::uuid,
     @title::text,
     sqlc.narg('mode')::chat_mode,
+    @status::chat_status,
     COALESCE(@mcp_server_ids::uuid[], '{}'::uuid[]),
     COALESCE(sqlc.narg('labels')::jsonb, '{}'::jsonb)
 )
@@ -550,6 +567,43 @@ WHERE
 RETURNING
     *;
 
+-- name: LinkChatFiles :one
+-- LinkChatFiles inserts file associations into the chat_file_links
+-- join table with deduplication (ON CONFLICT DO NOTHING). The INSERT
+-- is conditional: it only proceeds when the total number of links
+-- (existing + genuinely new) does not exceed max_file_links. Returns
+-- the number of genuinely new file IDs that were NOT inserted due to
+-- the cap. A return value of 0 means all files were linked (or were
+-- already linked). A positive value means the cap blocked that many
+-- new links.
+WITH current AS (
+    SELECT COUNT(*) AS cnt
+    FROM chat_file_links
+    WHERE chat_id = @chat_id::uuid
+),
+new_links AS (
+    SELECT @chat_id::uuid AS chat_id, unnest(@file_ids::uuid[]) AS file_id
+),
+genuinely_new AS (
+    SELECT nl.chat_id, nl.file_id
+    FROM new_links nl
+    WHERE NOT EXISTS (
+        SELECT 1 FROM chat_file_links cfl
+        WHERE cfl.chat_id = nl.chat_id AND cfl.file_id = nl.file_id
+    )
+),
+inserted AS (
+    INSERT INTO chat_file_links (chat_id, file_id)
+    SELECT gn.chat_id, gn.file_id
+    FROM genuinely_new gn, current c
+    WHERE c.cnt + (SELECT COUNT(*) FROM genuinely_new) <= @max_file_links::int
+    ON CONFLICT (chat_id, file_id) DO NOTHING
+    RETURNING file_id
+)
+SELECT
+    (SELECT COUNT(*)::int FROM genuinely_new) -
+    (SELECT COUNT(*)::int FROM inserted) AS rejected_new_files;
+
 -- name: AcquireChats :many
 -- Acquires up to @num_chats pending chats for processing. Uses SKIP LOCKED
 -- to prevent multiple replicas from acquiring the same chat.
@@ -620,17 +674,20 @@ WHERE
     status = 'running'::chat_status
     AND heartbeat_at < @stale_threshold::timestamptz;
 
--- name: UpdateChatHeartbeat :execrows
--- Bumps the heartbeat timestamp for a running chat so that other
--- replicas know the worker is still alive.
+-- name: UpdateChatHeartbeats :many
+-- Bumps the heartbeat timestamp for the given set of chat IDs,
+-- provided they are still running and owned by the specified
+-- worker. Returns the IDs that were actually updated so the
+-- caller can detect stolen or completed chats via set-difference.
 UPDATE
     chats
 SET
-    heartbeat_at = NOW()
+    heartbeat_at = @now::timestamptz
 WHERE
-    id = @id::uuid
+    id = ANY(@ids::uuid[])
     AND worker_id = @worker_id::uuid
-    AND status = 'running'::chat_status;
+    AND status = 'running'::chat_status
+RETURNING id;
 
 -- name: GetChatDiffStatusByChatID :one
 SELECT
@@ -866,7 +923,8 @@ SELECT
     COALESCE(SUM(cm.input_tokens), 0)::bigint AS total_input_tokens,
     COALESCE(SUM(cm.output_tokens), 0)::bigint AS total_output_tokens,
     COALESCE(SUM(cm.cache_read_tokens), 0)::bigint AS total_cache_read_tokens,
-    COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens
+    COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens,
+    COALESCE(SUM(cm.runtime_ms), 0)::bigint AS total_runtime_ms
 FROM
     chat_messages cm
 JOIN
@@ -896,7 +954,8 @@ SELECT
     COALESCE(SUM(cm.input_tokens), 0)::bigint AS total_input_tokens,
     COALESCE(SUM(cm.output_tokens), 0)::bigint AS total_output_tokens,
     COALESCE(SUM(cm.cache_read_tokens), 0)::bigint AS total_cache_read_tokens,
-    COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens
+    COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens,
+    COALESCE(SUM(cm.runtime_ms), 0)::bigint AS total_runtime_ms
 FROM
     chat_messages cm
 JOIN
@@ -931,7 +990,8 @@ WITH chat_costs AS (
         COALESCE(SUM(cm.input_tokens), 0)::bigint AS total_input_tokens,
         COALESCE(SUM(cm.output_tokens), 0)::bigint AS total_output_tokens,
         COALESCE(SUM(cm.cache_read_tokens), 0)::bigint AS total_cache_read_tokens,
-        COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens
+        COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens,
+        COALESCE(SUM(cm.runtime_ms), 0)::bigint AS total_runtime_ms
     FROM chat_messages cm
     JOIN chats c ON c.id = cm.chat_id
     WHERE c.owner_id = @owner_id::uuid
@@ -948,7 +1008,8 @@ SELECT
     cc.total_input_tokens,
     cc.total_output_tokens,
     cc.total_cache_read_tokens,
-    cc.total_cache_creation_tokens
+    cc.total_cache_creation_tokens,
+    cc.total_runtime_ms
 FROM chat_costs cc
 LEFT JOIN chats rc ON rc.id = cc.root_chat_id
 ORDER BY cc.total_cost_micros DESC;
@@ -974,7 +1035,8 @@ WITH chat_cost_users AS (
         COALESCE(SUM(cm.input_tokens), 0)::bigint AS total_input_tokens,
         COALESCE(SUM(cm.output_tokens), 0)::bigint AS total_output_tokens,
         COALESCE(SUM(cm.cache_read_tokens), 0)::bigint AS total_cache_read_tokens,
-        COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens
+        COALESCE(SUM(cm.cache_creation_tokens), 0)::bigint AS total_cache_creation_tokens,
+        COALESCE(SUM(cm.runtime_ms), 0)::bigint AS total_runtime_ms
     FROM
         chat_messages cm
     JOIN
@@ -1008,6 +1070,7 @@ SELECT
     total_output_tokens,
     total_cache_read_tokens,
     total_cache_creation_tokens,
+    total_runtime_ms,
     COUNT(*) OVER()::bigint AS total_count
 FROM
     chat_cost_users

coderd/database/queries/connectionlogs.sql

@@ -133,111 +133,113 @@ OFFSET
 	@offset_opt;
 
 -- name: CountConnectionLogs :one
-SELECT
-	COUNT(*) AS count
-FROM
-	connection_logs
-JOIN users AS workspace_owner ON
-	connection_logs.workspace_owner_id = workspace_owner.id
-LEFT JOIN users ON
-	connection_logs.user_id = users.id
-JOIN organizations ON
-	connection_logs.organization_id = organizations.id
-WHERE
-	-- Filter organization_id
-	CASE
-		WHEN @organization_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
-			connection_logs.organization_id = @organization_id
-		ELSE true
-	END
-	-- Filter by workspace owner username
-	AND CASE
-		WHEN @workspace_owner :: text != '' THEN
-			workspace_owner_id = (
-				SELECT id FROM users
-				WHERE lower(username) = lower(@workspace_owner) AND deleted = false
-			)
-		ELSE true
-	END
-	-- Filter by workspace_owner_id
-	AND CASE
-		WHEN @workspace_owner_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
-			workspace_owner_id = @workspace_owner_id
-		ELSE true
-	END
-	-- Filter by workspace_owner_email
-	AND CASE
-		WHEN @workspace_owner_email :: text != '' THEN
-			workspace_owner_id = (
-				SELECT id FROM users
-				WHERE email = @workspace_owner_email AND deleted = false
-			)
-		ELSE true
-	END
-	-- Filter by type
-	AND CASE
-		WHEN @type :: text != '' THEN
-			type = @type :: connection_type
-		ELSE true
-	END
-	-- Filter by user_id
-	AND CASE
-		WHEN @user_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
-			user_id = @user_id
-		ELSE true
-	END
-	-- Filter by username
-	AND CASE
-		WHEN @username :: text != '' THEN
-			user_id = (
-				SELECT id FROM users
-				WHERE lower(username) = lower(@username) AND deleted = false
-			)
-		ELSE true
-	END
-	-- Filter by user_email
-	AND CASE
-		WHEN @user_email :: text != '' THEN
-			users.email = @user_email
-		ELSE true
-	END
-	-- Filter by connected_after
-	AND CASE
-		WHEN @connected_after :: timestamp with time zone != '0001-01-01 00:00:00Z' THEN
-			connect_time >= @connected_after
-		ELSE true
-	END
-	-- Filter by connected_before
-	AND CASE
-		WHEN @connected_before :: timestamp with time zone != '0001-01-01 00:00:00Z' THEN
-			connect_time <= @connected_before
-		ELSE true
-	END
-	-- Filter by workspace_id
-	AND CASE
-		WHEN @workspace_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
-			connection_logs.workspace_id = @workspace_id
-		ELSE true
-	END
-	-- Filter by connection_id
-	AND CASE
-		WHEN @connection_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
-			connection_logs.connection_id = @connection_id
-		ELSE true
-	END
-	-- Filter by whether the session has a disconnect_time
-	AND CASE
-		WHEN @status :: text != '' THEN
-			((@status = 'ongoing' AND disconnect_time IS NULL) OR
-			(@status = 'completed' AND disconnect_time IS NOT NULL)) AND
-			-- Exclude web events, since we don't know their close time.
-			"type" NOT IN ('workspace_app', 'port_forwarding')
-		ELSE true
-	END
-	-- Authorize Filter clause will be injected below in
-	-- CountAuthorizedConnectionLogs
-	-- @authorize_filter
-;
+SELECT COUNT(*) AS count FROM (
+	SELECT 1
+	FROM
+		connection_logs
+	JOIN users AS workspace_owner ON
+		connection_logs.workspace_owner_id = workspace_owner.id
+	LEFT JOIN users ON
+		connection_logs.user_id = users.id
+	JOIN organizations ON
+		connection_logs.organization_id = organizations.id
+	WHERE
+		-- Filter organization_id
+		CASE
+			WHEN @organization_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
+				connection_logs.organization_id = @organization_id
+			ELSE true
+		END
+		-- Filter by workspace owner username
+		AND CASE
+			WHEN @workspace_owner :: text != '' THEN
+				workspace_owner_id = (
+					SELECT id FROM users
+					WHERE lower(username) = lower(@workspace_owner) AND deleted = false
+				)
+			ELSE true
+		END
+		-- Filter by workspace_owner_id
+		AND CASE
+			WHEN @workspace_owner_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
+				workspace_owner_id = @workspace_owner_id
+			ELSE true
+		END
+		-- Filter by workspace_owner_email
+		AND CASE
+			WHEN @workspace_owner_email :: text != '' THEN
+				workspace_owner_id = (
+					SELECT id FROM users
+					WHERE email = @workspace_owner_email AND deleted = false
+				)
+			ELSE true
+		END
+		-- Filter by type
+		AND CASE
+			WHEN @type :: text != '' THEN
+				type = @type :: connection_type
+			ELSE true
+		END
+		-- Filter by user_id
+		AND CASE
+			WHEN @user_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
+				user_id = @user_id
+			ELSE true
+		END
+		-- Filter by username
+		AND CASE
+			WHEN @username :: text != '' THEN
+				user_id = (
+					SELECT id FROM users
+					WHERE lower(username) = lower(@username) AND deleted = false
+				)
+			ELSE true
+		END
+		-- Filter by user_email
+		AND CASE
+			WHEN @user_email :: text != '' THEN
+				users.email = @user_email
+			ELSE true
+		END
+		-- Filter by connected_after
+		AND CASE
+			WHEN @connected_after :: timestamp with time zone != '0001-01-01 00:00:00Z' THEN
+				connect_time >= @connected_after
+			ELSE true
+		END
+		-- Filter by connected_before
+		AND CASE
+			WHEN @connected_before :: timestamp with time zone != '0001-01-01 00:00:00Z' THEN
+				connect_time <= @connected_before
+			ELSE true
+		END
+		-- Filter by workspace_id
+		AND CASE
+			WHEN @workspace_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
+				connection_logs.workspace_id = @workspace_id
+			ELSE true
+		END
+		-- Filter by connection_id
+		AND CASE
+			WHEN @connection_id :: uuid != '00000000-0000-0000-0000-000000000000'::uuid THEN
+				connection_logs.connection_id = @connection_id
+			ELSE true
+		END
+		-- Filter by whether the session has a disconnect_time
+		AND CASE
+			WHEN @status :: text != '' THEN
+				((@status = 'ongoing' AND disconnect_time IS NULL) OR
+				(@status = 'completed' AND disconnect_time IS NOT NULL)) AND
+				-- Exclude web events, since we don't know their close time.
+				"type" NOT IN ('workspace_app', 'port_forwarding')
+			ELSE true
+		END
+		-- Authorize Filter clause will be injected below in
+		-- CountAuthorizedConnectionLogs
+		-- @authorize_filter
+	-- NOTE: See the CountAuditLogs LIMIT note.
+	LIMIT NULLIF(@count_cap::int, 0) + 1
+) AS limited_count;
 
 -- name: DeleteOldConnectionLogs :execrows
 WITH old_logs AS (
@@ -251,55 +253,75 @@ DELETE FROM connection_logs
 USING old_logs
 WHERE connection_logs.id = old_logs.id;
 
--- name: UpsertConnectionLog :one
+-- name: BatchUpsertConnectionLogs :exec
 INSERT INTO connection_logs (
-	id,
-	connect_time,
-	organization_id,
-	workspace_owner_id,
-	workspace_id,
-	workspace_name,
-	agent_name,
-	type,
-	code,
-	ip,
-	user_agent,
-	user_id,
-	slug_or_port,
-	connection_id,
-	disconnect_reason,
-	disconnect_time
-) VALUES
-	($1, @time, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11, $12, $13, $14,
-	-- If we've only received a disconnect event, mark the event as immediately
-	-- closed.
-	 CASE
-		 WHEN @connection_status::connection_status = 'disconnected'
-		 THEN @time :: timestamp with time zone
-		 ELSE NULL
-	 END)
+    id, connect_time, organization_id, workspace_owner_id, workspace_id,
+    workspace_name, agent_name, type, code, ip, user_agent, user_id,
+    slug_or_port, connection_id, disconnect_reason, disconnect_time
+)
+SELECT
+    u.id,
+    u.connect_time,
+    u.organization_id,
+    u.workspace_owner_id,
+    u.workspace_id,
+    u.workspace_name,
+    u.agent_name,
+    u.type,
+    -- Use the validity flag to distinguish "no code" (NULL) from a
+    -- legitimate zero exit code.
+    CASE WHEN u.code_valid THEN u.code ELSE NULL END,
+    u.ip,
+    NULLIF(u.user_agent, ''),
+    NULLIF(u.user_id, '00000000-0000-0000-0000-000000000000'::uuid),
+    NULLIF(u.slug_or_port, ''),
+    NULLIF(u.connection_id, '00000000-0000-0000-0000-000000000000'::uuid),
+    NULLIF(u.disconnect_reason, ''),
+    NULLIF(u.disconnect_time, '0001-01-01 00:00:00Z'::timestamptz)
+FROM (
+    SELECT
+        unnest(sqlc.arg('id')::uuid[]) AS id,
+        unnest(sqlc.arg('connect_time')::timestamptz[]) AS connect_time,
+        unnest(sqlc.arg('organization_id')::uuid[]) AS organization_id,
+        unnest(sqlc.arg('workspace_owner_id')::uuid[]) AS workspace_owner_id,
+        unnest(sqlc.arg('workspace_id')::uuid[]) AS workspace_id,
+        unnest(sqlc.arg('workspace_name')::text[]) AS workspace_name,
+        unnest(sqlc.arg('agent_name')::text[]) AS agent_name,
+        unnest(sqlc.arg('type')::connection_type[]) AS type,
+        unnest(sqlc.arg('code')::int4[]) AS code,
+        unnest(sqlc.arg('code_valid')::bool[]) AS code_valid,
+        unnest(sqlc.arg('ip')::inet[]) AS ip,
+        unnest(sqlc.arg('user_agent')::text[]) AS user_agent,
+        unnest(sqlc.arg('user_id')::uuid[]) AS user_id,
+        unnest(sqlc.arg('slug_or_port')::text[]) AS slug_or_port,
+        unnest(sqlc.arg('connection_id')::uuid[]) AS connection_id,
+        unnest(sqlc.arg('disconnect_reason')::text[]) AS disconnect_reason,
+        unnest(sqlc.arg('disconnect_time')::timestamptz[]) AS disconnect_time
+) AS u
 ON CONFLICT (connection_id, workspace_id, agent_name)
 DO UPDATE SET
-	-- No-op if the connection is still open.
-	disconnect_time = CASE
-		WHEN @connection_status::connection_status = 'disconnected'
-		-- Can only be set once
-		AND connection_logs.disconnect_time IS NULL
-		THEN EXCLUDED.connect_time
-		ELSE connection_logs.disconnect_time
-	END,
-	disconnect_reason = CASE
-		WHEN @connection_status::connection_status = 'disconnected'
-		-- Can only be set once
-		AND connection_logs.disconnect_reason IS NULL
-		THEN EXCLUDED.disconnect_reason
-		ELSE connection_logs.disconnect_reason
-	END,
-	code = CASE
-		WHEN @connection_status::connection_status = 'disconnected'
-		-- Can only be set once
-		AND connection_logs.code IS NULL
-		THEN EXCLUDED.code
-		ELSE connection_logs.code
-	END
-RETURNING *;
+    -- Pick the earliest real connect_time. The zero sentinel
+    -- ('0001-01-01') means the batch didn't know the connect_time
+    -- (e.g. a pure disconnect event), so we keep the existing value.
+    connect_time = CASE
+        WHEN EXCLUDED.connect_time = '0001-01-01 00:00:00Z'::timestamptz
+        THEN connection_logs.connect_time
+        WHEN connection_logs.connect_time = '0001-01-01 00:00:00Z'::timestamptz
+        THEN EXCLUDED.connect_time
+        ELSE LEAST(connection_logs.connect_time, EXCLUDED.connect_time)
+    END,
+    disconnect_time = CASE
+        WHEN connection_logs.disconnect_time IS NULL
+        THEN EXCLUDED.disconnect_time
+        ELSE connection_logs.disconnect_time
+    END,
+    disconnect_reason = CASE
+        WHEN connection_logs.disconnect_reason IS NULL
+        THEN EXCLUDED.disconnect_reason
+        ELSE connection_logs.disconnect_reason
+    END,
+    code = CASE
+        WHEN connection_logs.code IS NULL
+        THEN EXCLUDED.code
+        ELSE connection_logs.code
+    END;

coderd/database/queries/tailnet.sql

@@ -96,28 +96,6 @@ DELETE
 FROM tailnet_tunnels
 WHERE coordinator_id = $1 and src_id = $2;
 
--- name: GetTailnetTunnelPeerIDs :many
-SELECT dst_id as peer_id, coordinator_id, updated_at
-FROM tailnet_tunnels
-WHERE tailnet_tunnels.src_id = $1
-UNION
-SELECT src_id as peer_id, coordinator_id, updated_at
-FROM tailnet_tunnels
-WHERE tailnet_tunnels.dst_id = $1;
-
--- name: GetTailnetTunnelPeerBindings :many
-SELECT id AS peer_id, coordinator_id, updated_at, node, status
-FROM tailnet_peers
-WHERE id IN (
-  SELECT dst_id as peer_id
-  FROM tailnet_tunnels
-  WHERE tailnet_tunnels.src_id = $1
-  UNION
-  SELECT src_id as peer_id
-  FROM tailnet_tunnels
-  WHERE tailnet_tunnels.dst_id = $1
-);
-
 -- For PG Coordinator HTMLDebug
 
 -- name: GetAllTailnetCoordinators :many
@@ -128,3 +106,22 @@ SELECT * FROM tailnet_peers;
 
 -- name: GetAllTailnetTunnels :many
 SELECT * FROM tailnet_tunnels;
+
+-- name: GetTailnetTunnelPeerIDsBatch :many
+SELECT src_id AS lookup_id, dst_id AS peer_id, coordinator_id, updated_at
+FROM tailnet_tunnels WHERE src_id = ANY(@ids :: uuid[])
+UNION ALL
+SELECT dst_id AS lookup_id, src_id AS peer_id, coordinator_id, updated_at
+FROM tailnet_tunnels WHERE dst_id = ANY(@ids :: uuid[]);
+
+-- name: GetTailnetTunnelPeerBindingsBatch :many
+SELECT tp.id AS peer_id, tp.coordinator_id, tp.updated_at, tp.node, tp.status,
+       tunnels.lookup_id
+FROM (
+  SELECT dst_id AS peer_id, src_id AS lookup_id
+  FROM tailnet_tunnels WHERE src_id = ANY(@ids :: uuid[])
+  UNION
+  SELECT src_id AS peer_id, dst_id AS lookup_id
+  FROM tailnet_tunnels WHERE dst_id = ANY(@ids :: uuid[])
+) tunnels
+INNER JOIN tailnet_peers tp ON tp.id = tunnels.peer_id;

coderd/database/queries/userchatproviderkeys.sql

@@ -0,0 +1,20 @@
+-- name: GetUserChatProviderKeys :many
+SELECT * FROM user_chat_provider_keys WHERE user_id = @user_id ORDER BY created_at ASC, id ASC;
+
+-- name: UpsertUserChatProviderKey :one
+INSERT INTO user_chat_provider_keys (user_id, chat_provider_id, api_key, api_key_key_id)
+VALUES (@user_id, @chat_provider_id, @api_key, sqlc.narg('api_key_key_id')::text)
+ON CONFLICT (user_id, chat_provider_id) DO UPDATE SET
+    api_key = @api_key,
+    api_key_key_id = sqlc.narg('api_key_key_id')::text,
+    updated_at = NOW()
+RETURNING *;
+
+-- name: UpdateUserChatProviderKey :one
+UPDATE user_chat_provider_keys
+SET api_key = @api_key, api_key_key_id = sqlc.narg('api_key_key_id')::text, updated_at = NOW()
+WHERE user_id = @user_id AND chat_provider_id = @chat_provider_id
+RETURNING *;
+
+-- name: DeleteUserChatProviderKey :exec
+DELETE FROM user_chat_provider_keys WHERE user_id = @user_id AND chat_provider_id = @chat_provider_id;

coderd/database/sqlc.yaml

@@ -247,6 +247,7 @@ sql:
           mcp_server_tool_snapshots: MCPServerToolSnapshots
           mcp_server_config_id: MCPServerConfigID
           mcp_server_ids: MCPServerIDs
+          max_file_links: MaxFileLinks
           icon_url: IconURL
           oauth2_client_id: OAuth2ClientID
           oauth2_client_secret: OAuth2ClientSecret