Commit Graph
1873 Commits
Author SHA1 Message Date
Steven Masley bc5dd20a05 feat: report Agent Relay builds in telemetry (#31407)
Telemetry builds now report whether Agent Relay created them. Before
this, relay workspaces looked like any other build.

A start build counts as a relay build when it has the
`agent_relay_session_id` parameter. Stop and delete builds never get the
field. For those builds, the new `agent_relay` field on `WorkspaceBuild`
reports:

- `provider`: `claude_code`, `cursor`, `ambiguous`, or `unknown`
- `client_platform`: Claude Code only. Lowercased, characters outside
`[a-z0-9_]` become `_`, and cut to 15 characters
- `attempt`: omitted when missing or not an integer, so `0` is sent
- `session_id_hash` and `pool_hash`: SHA-256 hashes

No other parameter values are sent. Other builds leave the field out.

The snapshot fetches these parameters with one new batch query. The
workspace-create event reads the stored parameters with one extra query,
only when telemetry is enabled, so it reports the same data as the
snapshot. `docs/admin/setup/telemetry.md` lists the parameters that are
read.

The collector side is
[coder/coder-telemetry-server#56](https://github.com/coder/coder-telemetry-server/pull/56).
Merge this PR first.

Refs
[RELAY-9](https://linear.app/codercom/issue/RELAY-9/telemetry-report-agent-relay-usage-from-codercoder)

<details>
<summary>Decision log</summary>

- **Detect relay builds from parameters.** This needs no relay changes.
Relay registration or
[CORE-426](https://linear.app/codercom/issue/CORE-426) provenance
metadata can come later.
- **Builds only, not workspaces.** `GetWorkspaces` doesn't return the
latest build's parameters, so a workspace field would cost another
query. The collector can join on `workspace_id`.
- **Provider comes from parameter names only.**
`agent_relay_claude_code_*` means Claude Code and `agent_relay_cursor_*`
means Cursor. A build with both reports `ambiguous`; a build with
neither reports `unknown`. coderd stores every declared parameter on
every build (ephemeral ones with their default), so a Claude Code start
after a stop still reports `claude_code`.
- **Start builds only.** Stop and delete builds of a relay workspace
carry the same parameters but are not dispatches, so the snapshot skips
them before querying parameters.
- **Hash session ID and pool.** Pool names are user content, and
sessions only need counting.
- **Known gaps.** Cursor has no client platform. Relay version and
health would need relay registration.
- **Parameter names are shared with coder/agent-relay.** If agent-relay
renames one, telemetry silently stops reading it.

</details>

<details>
<summary>Manual test on <code>develop.sh</code></summary>

Ran on `8ad7cd491f` with `CODER_TELEMETRY_ENABLE=true
CODER_TELEMETRY_URL=http://127.0.0.1:8081` and `go run
./scripts/telemetry-server --port 8081` as the collector.

Pushed three templates with no infrastructure, each declaring the same
parameters as the real modules:

- `relay-claude`: `agent_relay_session_id`, `_delivery_id`, `_pool`,
`_client_platform`, plus ephemeral `agent_relay_attempt` and
`agent_relay_claude_code_lock_to_account`
- `relay-cursor`: `agent_relay_session_id`, `_delivery_id`, `_pool`,
`agent_relay_cursor_pool_name`, `_idle_release_timeout`, `_repo_url`
- `relay-plain`: `region`

Created workspaces with `POST /api/v2/users/me/workspaces` and
`rich_parameter_values`, the same call the relay makes. Then ran stop
and start builds on the relay workspaces, and restarted coderd to get a
periodic snapshot.

| Workspace | Build 1 (create event and snapshot) | Builds 2 and 3
(stop, start) |
|---|---|---|
| Claude (`claude_in_slack`, attempt 1) | `claude_code`, platform,
attempt 1 | `claude_code`, platform, no attempt |
| Cursor | `cursor` | `cursor` |
| Plain | field omitted | n/a |
| Claude template, no session ID | field omitted | n/a |

- Stop and start builds keep `agent_relay_claude_code_lock_to_account`,
with an empty value, so Claude Code rebuilds still report `claude_code`.
- No session ID, pool name, or `region` value appears in the captured
telemetry. The hashes match `sha256` of the inputs.
- Fed the captured snapshot through the conversion code in
coder/coder-telemetry-server#56: all 8 builds converted, 6 of them with
`agent_relay` set.
- No telemetry errors or panics in the coderd logs.

This run predates the start-only change: build 2 (stop) carried the
field then. Stop and delete builds are now omitted, which
`TestTelemetry_WorkspaceBuildAgentRelay` covers.

</details>


---

*Generated by Coder Agents on behalf of @Emyrk.*
2026-11-03 15:48:10 -06:00
Ethan d70ce38559 feat: record the OAuth2 client registration method of MCP server configs (#31558)
Relates to CODAGT-779

Currently we can't tell whether the OAuth2 client of an MCP server
config was entered by an admin, or registered by Coder with Dynamic
Client Registration (RFC 7591). The next PR in the stack needs to know
this, so it can register a new client when a config is updated.

This PR adds `mcp_server_configs.oauth2_client_registration` (`manual`
or `dynamic`) and returns it to admins on the API. Create stores
`dynamic` when it registers the client, and `manual` otherwise. An
update resets it to `manual` when it switches away from the `oauth2`
auth type, or when it changes the client ID, secret, auth URL or token
URL of a `dynamic` config, since the admin has replaced Coder's client.

Existing rows default to `manual`, as there's nothing in the data that
tells us which clients came from Dynamic Client Registration.
2026-11-03 12:39:30 +11:00
55573833d0 fix(coderd/database): lock the workspaces row when claiming a prebuild (#31518)
`ClaimPrebuiltWorkspace` locked the `workspace_prebuilds` view with `FOR
UPDATE ... SKIP LOCKED`, but the view is built from CTEs, so nothing got
locked. Concurrent claims picked the same prebuild and the losers got a
409 (a customer saw 28 of 30 fail).

This locks the `workspaces` row directly (`FOR NO KEY UPDATE SKIP
LOCKED`), rechecks owner and deleted on it, and checks eligibility with
per-workspace index lookups instead of the views. It uses `=` instead of
`IN` so one claim can't take two prebuilds, filters on the preset's
template so other templates' prebuilds aren't scanned, and claims oldest
first so prebuilds are used before they hit the preset TTL. Query-only,
no migration, applies cleanly to release/2.38. Builds on #31468 by Kevin
Ha (Canva).

Benchmarks (Postgres 17, 20k users, 320k builds, 600 ready prebuilds):
~12 ms per claim (was ~640 ms), and 500 concurrent claims got distinct
prebuilds in ~1.5 s.

Known limitations:

- Each claim still checks every prebuild of the preset's template (other
presets included) before sorting. That's a few ms at hundreds of
prebuilds per template.
- Pre-existing: a reconciler delete during a claim can still 409.
Follow-up is a `deleting_at` marker the claim skips.

---------

Co-authored-by: Kevin Ha <kevinh@canva.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-30 08:43:52 -07:00
Mathias Fredriksson c0f7663e95 feat: support editing queued chat messages (#29215)
The owner of a chat can edit a queued message through
`PATCH /api/v2/chats/{chat}/queue/{queuedMessage}`. chatd never sends
a row under edit (`editing_since`): rows ahead of it still send in
order, and a turn that ends at it puts the chat in the new `paused`
status. Ending the head's edit, deleting it, or promoting a queued row
continues the chat. Interrupting sends still run ahead of the queue.

The chat state machine classifies a queue whose head is under edit as
its own sub-state (R1P, I1P, A1P, E1P, XE1P, P), so every path to
`paused` is in the transition matrix. A chat whose family has a paused
member cannot be archived, and auto-archive skips that family. Queue
edit and delete are owner-only, because both can send the next message
with the owner's credentials.

`ChatQueuedMessage` gains `editing_since` and `reasoning_effort`, and
`ChatStatus` gains `paused`.

Closes CODAGT-1028
Refs CODAGT-971
2026-10-30 15:52:22 +02:00
Michael Suchacz 31b8fd1b88 feat: graduate the chat advisor from experiments (#31414) 2026-10-29 23:24:18 +01:00
Michael Suchacz 1acacc9c30 feat: graduate MCP tool search from experiments (#31327) 2026-10-29 20:32:56 +01:00
Michael Suchacz d1f308d9f3 feat: let organization members turn off organization skills in chats (#31134)
## Stack

Organization skills load into every member's Coder Agents chats in their
organization, with no way for a member to opt out. This stack adds a
per-organization setting that lets members turn them off for their own
chats, bottom to top:

1. **#31134** data model, API, and chatd (this PR)
2. #31135 admin and member controls in the dashboard, the `/` menu, and
docs

## What changes

Organizations get `members_can_disable_skills`, off by default. Users
who can update organization skills change it with `PUT
/api/experimental/organizations/{organization}/skill-settings`, which
writes an organization audit entry. Anyone who can read the organization
can `GET` it.

While the setting is on, each member saves their own choices with `PATCH
/api/experimental/organizations/{organization}/members/{user}/skill-preferences`:
`all_skills_disabled` (a master switch that also covers skills created
later) and `skills`, a sparse map from skill ID to on or off. `GET`
returns the setting, the master switch, and `disabled_skill_ids`. The
routes follow the personal-settings rules: without `read_personal` or
`update_personal` on that user the caller gets 404, a PATCH while the
setting is off gets 403, and a key that is not an organization skill
that member can read gets a 400 that names the key. The check runs as
the member, not the caller, so an owner or user admin editing someone
else's choices is held to that member's skills. Turning the setting off
keeps the choices but ignores them.

Migration 000636 adds the column and two tables keyed by organization
membership. Both cascade when the membership goes away, and the
per-skill table also cascades when the skill is deleted, so a recreated
skill loads by default.

chatd reads the preferences as the chat owner when it builds the skill
index and when it loads a body for `read_skill` or for a structured
skill invocation (#30861). A skill the member turned off is left out of
the index, and both loads report it as not found, including later in a
running chat. If the preferences cannot be read, turn preparation fails,
as it already does when the skill metadata cannot be read.

Older migration tests created organizations with `dbgen.Organization`
and `GetDefaultOrganization`, which now select the new column before it
exists, so they insert organizations with raw SQL instead.

## Review focus

- Authorization in `coderd/organizationskillpreferences.go` and the five
new dbauthz methods (`read_personal` and `update_personal` on the user,
`organization_skill` update for the setting).
- `GetOrganizationSkillMemberPreferencesRow.Excludes` is the one rule
chatd applies; #31135 mirrors it in `activeOrganizationSkills`.

## Testing

- Go tests cover the queries and cascades, dbauthz, both routes
(including the audit diff and each 400 case), and chatd (index,
`read_skill`, skill invocations, master switch, a change during a chat).
Each new test fails with its code path removed.
- Remote dogfood UAT (10 scenarios, real model) passed on the pre-split
branch head `186c28d81ab`, which has the same code except the fix below.
It found one issue: two spellings of one skill ID with opposite values
stored a random result. The PATCH now rejects every aliased key with a
400. A second run on the stack's top head at the time, `718dd6d1bd1`,
passed: every aliased form returns 400 and stores nothing, and chats
leave out a skill the member turned off, including in `read_skill`.
- Polish and the test audit stayed inside the implementation diff. No
new dependencies.
- Rebased on main as one commit. The migration is now 000636, after
main's 000634 and 000635. Every organization skill load passes the chat
owner, including main's skill invocation path, so `/skill` invocations
honor the opt-out too. The model's skill index leaves out a skill that
is not model-invocable or that the member turned off. Apart from that,
the diff is unchanged.

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-29 12:53:50 +00:00
Michael Suchacz 3697bb38ce feat!: choose the computer use model per organization (#31300)
> [!NOTE]
> Xum opened this PR for Mike (Coder agent). Mike reviews it before
merge.

Fixes CODAGT-1472.

Computer use now runs on a chat model that each organization picks.
Before, it ran a fixed model on the first enabled AI provider of a
deployment-wide provider type, sorted by name. On dogfood that
provider's key is rejected, so every computer use subagent failed.

## What changed

- **Setting.** A new `computer_use` organization model override, set in
the **Computer use** row of **Organization settings** on the Coder
Agents admin page. Without an override, computer use runs on the
organization's default model. If the chosen model is disabled, deleted,
or unusable for the chat owner, computer use falls back to the default
model, as other overrides do. It also falls back when the chosen model
moves to an AI provider that is not Anthropic or OpenAI, and the row
then says that the saved model is ignored.
- **API.** `PUT
/api/v2/organizations/{organization}/chats/model-overrides/computer_use`
returns 404 while the `chat-virtual-desktop` experiment is off, and 400
unless the model's AI provider type is Anthropic or OpenAI.
- **chatd.** The `computer_use` subagent runs on that model, and the
computer tool flavor follows the model's provider type. OpenAI computer
use always uses the Responses API, so a model that sets
`use_responses_api: false` gets a configuration error. When no usable
model exists, `computer_use` is unavailable, and the reason says whether
a saved computer use model was skipped and why the default model cannot
be used. The default model is read under the chat owner's model ACLs, as
chat turns do.
- **UI.** The row lists only enabled Anthropic and OpenAI models. A
warning icon with the tooltip "Make sure that the selected model
supports computer use." appears when the model in effect (the selected
one, or the default) is not on a curated list of canonical model IDs
known to accept the computer tool (`knownComputerUseModels` in
`knownModels/catalog.ts`, exact match only). It also appears when the
model's AI provider type is OpenAI and its **Use Responses API**
override is Off, because OpenAI computer use needs the Responses API.
The warning does not block saving. I built the list from a live probe on
2026-10-26 that sent each Anthropic and OpenAI model one request with
the computer tool.
- **Docs.** `virtual-desktop.md`, `models.md`, `organizations.md`, and
`architecture.md` describe the new setting, and chatd's
`ARCHITECTURE.md` describes the computer use model routing.

![The Computer use row with an uncurated model selected shows the
warning icon and its
tooltip](https://github.com/user-attachments/assets/3948d936-5d8f-4ff9-a6de-14094c154dc8)

![The Computer use model picker lists only Anthropic and OpenAI
models](https://github.com/user-attachments/assets/913cd4a0-3cdf-4e5d-b2f9-582b9fca999c)

## Breaking change

This removes the deployment-wide computer use provider setting
(`agents_computer_use_provider`), its experimental `GET`/`PUT
/api/experimental/chats/config/computer-use-provider` endpoints (shipped
since v2.35), the Virtual desktop settings card, and its audit field.
Migration 000635 deletes the stored value. After the upgrade, computer
use runs on each organization's default model until an admin picks a
computer use model. Computer use subagents created before the upgrade
keep the model they inherited from their parent chat, so a follow-up
message to one of them can fail; a new computer use subagent uses the
organization's computer use model. Telemetry now reports the
per-organization overrides instead of the provider, and `null` when it
cannot read them.

## Testing

- `make pre-commit` passes. `go test` passes for `./coderd/x/chatd/...`,
`./coderd/database/migrations/`, and the targeted coderd, telemetry,
dbauthz, enterprise audit, and codersdk tests. Vitest and the touched
Storybook stories pass.
- Each new Go test fails when its target behavior is removed from
production code.
- Remote dogfood UAT ran on an isolated Coder instance built from the PR
head, with real Anthropic and OpenAI models:
- 110234e: the row with the experiment on and off, the picker, the
warning icon cases, save and clear, the 400 for other provider types,
which model and computer tool each computer use subagent used
(Anthropic, OpenAI, default, and disabled override), a real desktop
session on an Anthropic and an OpenAI model, the unavailable reasons,
the upgrade through migration 000634, and the removed endpoints.
- 09ce818: an override whose model moved to an unsupported AI provider.
- 7925be9: the alert for a moved and a disabled override, and a computer
use subagent on an Anthropic model.
- 2a01eaf: the row, and computer use subagents on the override and on
the default model.
- 47f5bd1: the warning icon for an OpenAI model with **Use Responses
API** set to Off, On, and Default. With Off, a computer use subagent on
that model fails with a configuration error before any provider request.
With Default, it runs with the computer tool through the Responses API.
- 81d33b4: the unavailable reasons for a disabled override, a cleared
override, and missing credentials, and a computer use subagent on an
Anthropic override. This round found that the docs named the wrong
section of the model form for **Use Responses API**. aeb3d02 fixes that
step, and a re-read of the docs against the recorded UI passed.
2026-10-29 13:00:05 +01:00
Sas Swart eb38759fda feat: limit effective group resolution to an organization (#31119)
AI budget resolution searches all of a user's groups across every
organization. Whichever group wins determines the organization. #31120
needs to resolve inside a known organization, and it needs to know which
organization site-wide resolution picked.

Add an `organizationID uuid.NullUUID` argument to
`budget.ResolveUserAIBudget` and `budget.ResolveUserEffectiveGroup`.
When it is set, count only that organization's override, group budgets
and Everyone group. An organization that is unknown, deleted or that the
user is not a member of resolves to nothing. When it is null, resolve
exactly as before. Return the winning group's organization in
`budget.EffectiveGroup.OrganizationID`.

Read the override through a new resolution-only query,
`GetUserAIBudgetOverrideWithOrganization`, so that the CRUD query
`GetUserAIBudgetOverride` and the admin endpoints stay untouched. Add a
nullable `organization_id` filter to `GetHighestGroupAIBudgetByUser` and
`GetUserEveryoneFallbackGroup`.

Every caller passes null, so behaviour is unchanged. There is no
migration and no proto change.

Scoping to an organization can lift a limit: a user capped in org B has
no limit in org A if A has no budgets. AIGOV-711 intends this, and it
takes effect in #31120.

Part of AIGOV-715.

---

*This PR was generated by Coder Agents on behalf of @SasSwart.*
2026-10-29 12:14:34 +02:00
Michael Suchacz 7f6ce9c8ab test(coderd): stamp chat message fixtures with the current content version (#30996)
Chat prompt tests built most of their stored-message fixtures in the
legacy content format (`content_version` 0), so they exercised the
legacy decoder instead of the current one. This PR changes tests only:
it stamps those fixtures with `ContentVersionV1` and builds them with
`MarshalParts`, so they keep passing after the legacy decoder is removed
later in this stack. Tests of the version 0 decoder itself keep their
version 0 rows until the layers above remove that decoder.

**Why it is safe**

- `mustMarshalContent` and `mustMarshalToolResult` now build current
parts (`PartFromContent`, `ChatMessageToolResult`) and encode them with
`MarshalParts`, the production write path.
- `TestMixedFormatConversation` becomes
`TestConvertMessagesWithFiles_Conversation` and uses current-format rows
only.
- `TestParseContent_BackwardCompat` does not change. Migration 000434
stamped every existing row with version 0, so its six cases of
current-format parts in version 0 rows still cover the version 0
heuristics. #30998 moves those six cases to version 1 rows when it
removes that decoder.
- `TestProviderExecutedResult_LegacyToolRow` calls the legacy writers
(`MarshalContent`, `MarshalToolResult`) directly, so its
provider-executed tool row still goes through the legacy tool-row
decoder.
- `TestParseContent_ErrorPaths` runs its malformed-input cases on
version 1 rows, and runs the same input on version 0 rows of each role
(`v0/<role>/malformed_json`), so a version 0 parser that accepts
malformed content still fails the suite. #30998 deletes the version 0
cases with that decoder.
- I ran the 17 mutants listed below, each a `go test -overlay` edit of
the current decode and prompt-conversion code in `chatprompt.go`,
against the `chatprompt` and `db2sdk` tests on `main` and on this
branch. Every mutant `main` catches is still caught. Two more are caught
here: mutant 3, because
`TestParseContent_ErrorPaths/system/malformed_json` used to hit the
legacy system decoder, and mutant 14, by the new JSON `null` case of
`TestConvertMessagesWithFiles_NormalizesAssistantToolCallInput`. Mutants
4 and 8 survive those tests on both. The `chatd` test
`TestOrganizationSkillsFollowChatOwnerAccess` catches mutant 4, so 8 is
an existing gap.
- Mutants of the version 0 decoder fail on `main` and on this branch:
version 0 assistant or user rows always take the envelope path, their
current-format parts are dropped, legacy tool rows lose
`provider_executed`, or a version 0 parser accepts malformed JSON.
- With the version 0 case removed from `ParseContent` by an overlay,
only tests of version 0 rows fail. #30997 deletes them, except the six
`TestParseContent_BackwardCompat` cases, which #30998 moves, and the
version 0 cases of `TestParseContent_ErrorPaths`, which #30998 deletes.

<details>
<summary>The 17 mutants</summary>

1. `ParseContent` returns no parts for a V1 row.
2. V1 decoding skips NUL restoration (`decodeNulInParts`).
3. The V1 parse error says "decode" instead of "parse".
4. Tool error string extraction is off (`toolResultPartToMessagePart`).
Only a `chatd` test catches it.
5. Tool call input is dropped.
6. Tool call `ProviderExecuted` is dropped when reading.
7. Resolved file data is dropped.
8. The file media type fallback is dropped (`partsToMessageParts`).
Survives.
9. User file IDs are not collected for resolution.
10. Text provider metadata is dropped.
11. File reference text is replaced by the raw content.
12. System messages are dropped.
13. Tool call `ProviderExecuted` is dropped when writing
(`PartFromContent`).
14. `normalizeToolCallInput` keeps a JSON `null` input instead of `{}`.
Survives on `main` only.
15. Text tool result `ProviderExecuted` is dropped.
16. Tool result IDs are dropped.
17. Assistant file IDs are resolved like user file IDs.

</details>

**Review focus**

The rewritten conversation test in `chatprompt_test.go`. It checks the
same prompt shape as before, but every row is in the current format. In
the same file, `TestProviderExecutedResult_LegacyToolRow` keeps its
legacy rows, `TestParseContent_ErrorPaths` checks malformed input on
both content versions, and the input normalization table gains a JSON
`null` case. The other changed test files only switch fixtures to the
current format: `coderd/x/chatd/chatprompt/file_inline_test.go`
(+4/-16), `coderd/database/db2sdk/db2sdk_test.go` (+11/-13),
`coderd/x/chatd/message_conversion_test.go` (+3/-2) and
`coderd/x/chatd/quickgen_internal_test.go` (+1).

**Validation**

- `go test ./coderd/x/chatd/... ./coderd/database/db2sdk/ -count=1 -p 4
-parallel 4`: ok
- `go vet` on the touched packages and `golangci-lint run --fix=false
./coderd/x/chatd/chatprompt/...`: clean
- pre-commit hook: passed

**Stack**

This stack retires chatd compatibility paths for chat data that older
Coder versions stored.

1. #30996 stamp chat message fixtures with the current content version
(this PR)
2. #30997 drop tests of the legacy content codec
3. #30998 retire the V0 chat message content decoder
4. #31000 inline the model config provider identity
5. #31005 retire the `close_agent` tool alias

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-28 19:45:29 +00:00
Sas Swart 161b7233d9 feat(coderd): store the organization on AI Gateway interceptions (#31118)
Adds `organization_id` (FK to `organizations`, `ON DELETE RESTRICT`) and
`organization_source` (new enum: `workspace`, `chatd`,
`effective_group`) to `aibridge_interceptions`, with a check constraint
that keeps them set or unset together. Cost and pricing run later, in
`RecordTokenUsage`, so they need the organization that AIGOV-716a
resolves once per request stored on the interception row.

`RecordInterceptionRequest` gains `organization_id` (18) and
`organization_source` (19), and the aibridged API moves to v1.5.
`RecordInterception` stores a valid pair; any other combination is
logged as a warning and stored as NULLs, so a gateway bug never turns
into an HTTP 500 for the user. A well-formed ID for an organization that
does not exist fails on the FK, like an unknown `initiator_id`. Nothing
sends the fields yet, so behaviour is unchanged until AIGOV-716a.

**Malformed values are stored as empty, not rejected.** A failed
`RecordInterception` returns HTTP 500 to the user, so the server logs a
warning and stores NULLs instead. The one exception is a well-formed ID
for an organization that does not exist: the foreign key rejects it, the
same as an unknown `initiator_id` today.

Part of AIGOV-712

*This PR was generated by Coder Agents on behalf of @SasSwart.*
2026-10-28 10:52:35 +02:00
Michael Suchacz 7283fff69f fix(coderd/database/pubsub): close listener before waiting on latency probe (#31317) 2026-10-27 19:31:36 +01:00
Hugo Dutka 1a9d5b28ba fix: keep ACP harness context synchronized (#31351)
ACP harness resources bypass context drift detection, but only MCP
resources were live-synced into bound chats. Later ACP discoveries,
capability updates, probe errors, and removals could therefore leave
chat context stale without offering a refresh. Extend the existing
live-sync query to include ACP harnesses and publish context updates for
affected chats while keeping instructions and skills pinned.

Depends on #31350. This PR contains the ongoing synchronization change;
#31350 fixes initial probe ordering.
2026-10-27 18:28:16 +00:00
Ehab Younes 0d1cba2def feat: expose connection duration and client session ID in connection logs (#30759)
Show how long each SSH and reconnecting PTY connection lasted, and let
users find connections that share a client session ID. No schema changes.

API (ssh_info):
- Add client_session_id, the ID supplied by the client (the Coder CLI
  or a tool that launches it), if any.
- Add a client_session_id:<32 hex chars> filter.

UI:
- Each row shows the connect → disconnect time range and the duration
  as badges. A connection with no disconnect shows "Ongoing", equal
  timestamps show an "Unknown" start, and a disconnect recorded before
  its connect is labeled as such.
- The info icon opens a details popover instead of a hover tooltip. It
  uses the default popover styling and sentence-case labels, and lets
  users copy the client session ID and open "View related connections"
  for the same workspace.

Closes CLIENT-1010
2026-10-27 14:04:37 +03:00
Ehab Younes 28192213ab feat!: label the VS Code family and report built-in app usage per app (#29906)
The session banner now labels the VS Code family "VS Code Family", so it isn't confused with the VS Code editor. coderd_insights_applications_usage_seconds now reports built-in apps per app instead of per family.

- Generate the frontend's family labels from the Go table in codersdk.
- Add the connection_method label: workspace_app for template apps, and ssh, reconnecting_pty, or port_forwarding for built-in apps.
- Rename the application_name label to display_name. For built-in apps it holds the registry name, "Unknown" for unknown, and "Other apps" for overflow.
- Built-in apps get one series per app name, only for apps used in the window. The series limit and the fold rule are shared with the template insights API.
- Breaking: selectors with a value for application_name, or with slug="", match nothing. The Prometheus docs map each old series to its new one.

Closes #29940
Closes CLIENT-1034
Closes CLIENT-1036
2026-10-27 09:58:36 +00:00
Michael Suchacz 3f055e7a56 feat: list every chat attachment in the Summary panel (#31140)
In long chats, the Summary panel's **Attachments** section only listed
agent files from the messages already loaded in the transcript. It
showed "No attachments in the loaded messages." and a **Load earlier
messages** button, so finding a chat's files meant paging through its
history.

## What changes

- **API:** `GET /api/v2/chats/{chat}/attachments` returns every file
part in the chat's user-visible user and assistant messages, newest
first, as `{message_id, part_index, role, file}`, where `role` is `user`
or `assistant`. An optional `before_id` limits it to older messages.
Parts go through the same conversion as `GET /messages`, so `part_index`
matches that endpoint. `file` is a `ChatAttachmentFile`, a file-only
type, so the OpenAPI schema, the API reference, and the generated
TypeScript all describe a file part. An inline file omits `file_id` and
carries its bytes in `data` as a base64 string. It needs the same
permission as reading the chat. A new sqlc query,
`GetChatMessagesWithFilePartsByChatID`, backs it; there is no migration.
- **Panel:** attachments from the loaded messages stay live. While older
history is not loaded, the panel fetches the attachments before the
oldest loaded message and lists them with the rest. It sends no request
when every message is loaded or the Summary tab is hidden. Loading
earlier transcript pages reuses the list it already fetched instead of
sending another request, and does not blank, duplicate, or reorder it,
including while the last page moves into the chat store.
- **Groups:** files the agent attached are listed under "From the
agent", then uploads and pasted text under "From you".
- **Cards:** selecting a card previews the file. Images open in the
image lightbox and show their thumbnail on the card. Text, Markdown,
CSV, and JSON files open in the text preview, which opens at once with a
spinner while a stored file's text loads on its first view, and reuses
the text until the cache frees it, five minutes after nothing on the
page reads it. An icon on the card downloads the file. Other files keep
the download card. Closing a preview returns focus to its card, and an
image whose thumbnail fails to load becomes a download card that takes
focus, even if its preview was open.
- **Dialogs:** `Lightbox` now passes Radix's close event to
`onCloseAutoFocus` instead of always calling `preventDefault`, so a card
hands it to `useDialogFocusReturn().restoreFocus`, which prevents the
default only when it can focus a target. `MermaidDiagram`, the only
other caller that passes the prop, now calls `preventDefault` itself, so
its focus return is unchanged. `ImageLightbox` and `TextPreviewDialog`
take the same optional prop.
- **Transcript focus:** closing the preview of an attachment in the
transcript (the image lightbox or the text preview) or of a tool result
image returns keyboard focus to the attachment that opened it instead of
the page body. Each attachment button passes itself to the opener,
because a stored text file opens its preview only after its download
finishes, when the click event no longer holds the button.
- **States:** the header count shows once the list is complete and
leaves out expired files. An empty list says "No attachments yet." While
the earlier request is pending, the loaded attachments stay, with a
spinner below them. If the request fails, they stay and the section
shows the server's error in an alert with Retry, which stays while the
retry runs or waits offline. The load-earlier button is gone.
- **Expired files:** files the chat no longer has fold into a closed "N
expired attachments" list that names each one and explains why files
expire. A file from unloaded history counts as expired when the chat's
file list no longer has it. If the cached chat record is older than
those messages (a revisit after more than a page of new messages), a new
file can show as expired until the chat refetch returns. A stored text
file whose preview finds it deleted stays in place as "Attachment
expired", keeps focus, and leaves the count.

| Cards grouped by who attached them | Earlier attachments in the
Summary panel | Expired files, listed by name |
| --- | --- | --- |
| ![Agent files under From the agent and an upload under From you, with
an image thumbnail and a download icon on each
card](https://github.com/user-attachments/assets/02683b78-5547-409e-ad02-d9a28eb8787c)
| ![Summary panel listing a file from a loaded message and two files
from messages that are not
loaded](https://github.com/user-attachments/assets/5e8c8916-a5ff-4af4-be62-1048ff557461)
| ![Opened 3 expired attachments list naming each expired file under an
explanation of why files
expire](https://github.com/user-attachments/assets/88ba6b37-4fc6-489e-9347-28b4b7dbe79f)
|

| Text file found deleted | Earlier request failed | Image preview |
| --- | --- | --- |
| ![PLAN.md card in place with a warning tile and Attachment expired,
and the header count down to
5](https://github.com/user-attachments/assets/7005bf19-f889-46a6-9e2a-78daa6c3be0c)
| ![A loaded attachment still listed above an alert with the server's
error and a Retry
button](https://github.com/user-attachments/assets/0348c762-5042-4b1d-97bc-64328334f4b0)
| ![Image lightbox open over the panel after selecting the image
card](https://github.com/user-attachments/assets/4af74d45-b1fc-4f2f-a275-0eb0c00da977)
|

Closes CODAGT-1436.

## Testing

`TestGetChatAttachments` checks the order and the filtering (deleted,
model-only, and tool messages are left out, and `part_index` skips a
stripped hook-context part as `GET /messages` does), compares every
returned file with its part from `GET /messages`, including an inline
paste's `data`, and covers `before_id`, a 400 for an invalid
`before_id`, a 404 for another user, and the raw JSON of an inline file,
which has no `file_id` key and a base64 `data` string. The dbauthz suite
asserts chat read. Vitest covers the panel's request gating and Retry,
seeding an earlier boundary from a cached list, the attachment helpers
(including earlier files the chat no longer lists), and the cards'
previews (including a text preview that opens before its text loads),
download link (including the share sheet it opens in an installed iOS
app), and focus, including a thumbnail that fails to load, also with its
preview open, and a text file that returns 404, and that a viewed text
file's text leaves the cache once nothing shows it. Vitest also opens a
transcript image, an inline text file, a stored text file whose download
finishes after focus has moved on, and a tool result image with Enter,
closes each with Escape, and checks that focus is back on the
attachment. Stories cover the new states, including a retry in flight, a
retry paused offline, and a reconnect after a failed retry. Two existing
stories now import the shared 1x1 PNG fixture instead of keeping their
own copies. Each new test fails when its code path is broken. No new
dependencies.

## Review record

- mafredri's review: grouping by role, the expired list, and keeping the
loaded attachments visible while the earlier request loads are in
`5cf0757b1d8`. The previews and the download icon that @ibetitsmike
asked for in the same round are in `6ae8719d6a0`.
- Frontend review (8 reviewer roles): 4 rounds before the first UAT, 3
delta rounds on the Codex fixes, then 12 rounds on mafredri's review and
the preview cards, with an FE1 to FE10 audit in rounds 2 to 10 (rounds
11 and 12 re-checked comments only), then 3 delta rounds on the
download-link test in `868d42430f0`, 4 on the fixes for Codex's fifth
run, 1 on its sixth, 7 on its seventh, and 4 on its eighth. The last of
those four passed every role it ran, and so did its FE1 to FE10 audit.
Every blocking and should-fix finding is fixed or declined with a
reason. Declined: the alert's title is an h2 under the section's h3,
because the shared `AlertTitle` always renders an h2, as in every
`ErrorAlert`; Retry does not move focus after a successful retry removes
it, which matches the other `RetryButton` callers and would need a focus
effect; the endpoint and panel ship together because the panel is the
endpoint's only consumer; a chat record without `files` is not read as
an empty file list, because the server also omits `files` when its
best-effort metadata query fails, which would show every earlier file as
expired; the header count does not subtract a deleted image or an
unopened text file, because the panel learns of a deletion only from the
chat record or a preview's 404; a card whose file a chat refetch reports
deleted moves into the expired list, which closes its open preview, as
the transcript swaps an evicted file's card for the expired tile; if the
transcript's last page arrives while the earlier request is still
loading, the header count briefly leaves out that page's files, because
only a rerender test reading the header could hold a fix, which FE1
keeps out of Vitest; a text file found deleted counts again once the
cache frees its 404, five minutes after nothing on the page reads it,
because keeping that 404 until reload would need a cache entry per
deleted file that is never freed, and the transcript on `main` forgets
the same 404 as soon as the chat closes; and AttachmentBlocks.tsx is at
719 lines (699 on `main`), because moving its file cards out would
create an import cycle.
- Remote dogfood UAT passed on `c1af9dd7cbc` and again on `d11c699cd0d`.
The second round ran `d11c699cd0d` and the `c1af9dd7cbc` frontend
against the same backend: an evicted earlier upload shows as expired on
`d11c699cd0d` and as a download card that returns 404 on `c1af9dd7cbc`,
and loading earlier pages sends no further attachments requests on
`d11c699cd0d`, while `c1af9dd7cbc` sent one per new page boundary. One
issue also reproduces on `main` and is unchanged here: after a live
attach evicts an older file, the panel and the transcript keep its
download card until the page reloads. The changes since `66e92f09064`
(mafredri's review, the preview cards, and the fixes for Codex's third
to eighth runs) were checked in Storybook and Vitest, and on dogfood
only through the keyboard round below, which also opened the Summary
panel's image and text cards; the screenshots above are Storybook
renders.
- Codex review: round 1 raised 4 findings, all fixed (see the resolved
threads). Round 2 found no issues on `d11c699cd0d`. A third run on
`d11c699cd0d` raised 2 findings, both fixed in `db0109d1120`. A fourth
run on `6ae8719d6a0` raised 1 finding (the download-link test read only
the link's attributes), fixed in `868d42430f0`. A fifth run on
`868d42430f0` raised 4 findings. Three are fixed in `e924d0595d9`: an
open image preview lost focus when its thumbnail failed, a slow text
load let a second preview open, and the earlier attachments dropped for
a render when the last page loaded. The fourth (paginate or cap the
endpoint) is left open for a product decision. A sixth run on
`e924d0595d9` raised 1 finding (the expired list's button did not say
what expired), fixed in `661db9b52af`. A seventh run on `661db9b52af`
raised 4 findings, all fixed: an inline file's `file_id` came back as
`null` and the schema described `data` as integers (`5fb4a047e39`), and
viewed text stayed cached until reload and the card's doc comment
restated its code (`51c5c98feb8`). An eighth run on `51c5c98feb8` raised
3 findings. Two are fixed: the attachment role allowed `system` and
`tool` (`2ed84cd882f`), and the error line hid the server's message
(`ce4e79593f6`). The third, that Testing Library cannot find the
decorative thumbnail by its `img` role, does not reproduce and is
declined in its thread.
- @jaaydenh's approval note, that keyboard focus was lost after closing
the preview of an attachment in the transcript, is fixed in
`b37b9952898` and `484e010dd25`; `359dd039b78`, `277b89ba1e2` and
`606b601c445` only refactor the hook, tighten the tests, and reword a
comment. Frontend review ran 5 rounds on the fix; the last passed, and
every blocking and should-fix finding is fixed or declined. Declined:
the composer's pending-attachment previews, the recording preview, and
the Instructions page's text preview also drop focus when they close,
but they predate this PR and are outside the transcript, so they are a
follow-up; if a stored text file finishes downloading while the same
message's image preview is open, closing the image returns focus to the
text file, because a message row remembers one opener, so focus still
stays in that message; and the fix grows ConversationTimeline.tsx from
812 to 827 lines, which was already over the 500-line guideline. Remote
dogfood UAT passed on `484e010dd25` with real key presses: after Enter
or Space opens the preview of a user's image, text file (also when its
download is slowed and the user presses Tab while it loads), or pasted
text, of an assistant's attachments, of one of three attachments in one
message, or of a tool result image, Escape returns focus to the same
attachment, so Tab continues from there. On `bac7cb055ce` focus fell to
the page body and the next Tab went to the Coder logo, so a keyboard
user needed 130 Tabs to get back to the composer instead of 44. Summary
panel cards returned focus before the fix and still do. The image,
slowed text file, and Summary panel cases passed again on `277b89ba1e2`;
`606b601c445` changes only a comment.
- Rebased onto `main` after those reviews. The added and removed lines
are unchanged; only the path of `AgentChatPageView.tsx` changed, because
`main` moved it into `AgentChatPage/`. After the eighth Codex run,
`main` was merged in, with one conflict in `coderd/chat_routes.go`,
where both sides added a route. On 2026-10-26 the branch was rebased
onto `main` (`aa9b0a950ad`) again, with conflicts in the same route list
and in the generated `site/src/api/typesGenerated.ts`, where both sides
added a type; the added and removed lines are unchanged.

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-27 03:13:52 +01:00
dylanhuff-at-coder 1e0a481a60 fix: retry workspace quota commits with jittered backoff (#31207)
Bursts of builds for the same user and org (usually autostarts) can hit
serialization failures in `CommitQuota`. `InTx` only retried those 3
times back-to-back, so once retries ran out the build failed outright.

`CommitQuota` stays `SERIALIZABLE` and now gets up to 10 retries with
full-jitter exponential backoff (50ms base, 2s cap, about 11s of backoff
at most). These builds aren't interactive, so a longer wait is a better
trade than a failed build. The wait respects the request context and
happens between transactions, so nothing holds a connection while it
sleeps. Each attempt also resets its results now, so a rolled-back
attempt can't leak `Ok: true` into the response.

The retry settings are opt-in fields on `database.TxOptions`
(`SerialMaxAttempts` and `BeforeSerialRetry`). Every other `InTx` caller
keeps the default 3 attempts with no delay. Since this touches
`coderd/database/db.go`, it'll need a look from a `coderd/database`
owner.

This replaces #30126 with the simpler approach Spike and I landed on. If
we still see failures after this ships, the next step is making
autostart replica-aware so a single replica handles each user, rather
than adding locking here.

Fixes [CORE-198](https://linear.app/codercom/issue/CORE-198)
2026-10-26 15:38:48 -07:00
TJ 4e4a56b665 fix: show configured provider icons and names in AI settings (#31033)
Admin-configured provider icons were ignored on the AI Settings Models
page (model rows, Add model menu, provider filter, model form header and
provider select) and in Agents compaction settings. These views derived
the icon only from the provider type, so a provider like DeepSeek on an
OpenAI-compatible type showed a generic icon even though the Providers
page and chat model selector showed the configured one. They now pass
the configured icon to `ProviderIcon`, which falls back to the type
icon.

The Spend page showed provider types (OpenAI) instead of configured
providers (DeepSeek) because `ListOrganizationAISpendUsers` aggregated
`ai.provider` while its filter matched `ai.provider_name`. It now
aggregates `provider_name`, and the page resolves names to display names
and icons from the AI Gateway providers list, which it shares with the
provider filter through one query. `OrganizationAISpendUser.providers`
now holds provider names; records from before configured names existed
have the type as their name, so they render as before.

AI Gateway Sessions views have the same type-versus-name issue and are
left for a separate fix.

**Note:** the Spend page resolves provider display names and icons from
`/aibridge/providers`, which requires deployment-wide read access
(`viewAnyAIBridgeInterception`). Owners and Auditors see the custom
display name and icon; Template Admins, User Admins, Org Admins, Org
Auditors and Org User Admins see the provider name rendered as a type
without the icon. Revisiting that endpoint's permissions is tracked in
AIGOV-726.

Refs CODAGT-1421

<details>
<summary>Decision log</summary>

- Root cause: callers rendered `<ProviderIcon provider={type} />`
without the descriptor's `icon`; nothing was stale.
- Spend page provider metadata is fetched only for viewers who can
already use the provider filter (`viewAnyAIBridgeInterception`); other
viewers see names formatted like provider types.
- Extended `ProvidersBadge` with an optional `providersByName` map
instead of adding a Spend-specific badge, so the Sessions fix can reuse
it.
- Removed `providerTypeByIDFromUserConfigs`, which became unused once
compaction settings used `providerInfoByIDFromUserConfigs`.
- Visual coverage is in Storybook stories (ModelRow, ModelForm,
UserCompactionThresholdSettings, SpendPage) per the repo's FE1 rule.

</details>

---

_Generated by Coder Agents on behalf of @tracyjohnsonux._
2026-10-26 13:26:38 +00:00
Susana Ferreira 15a5844253 feat: add cost to AI Gateway session threads API (#30509)
## Description

Exposes cost data on the AI Gateway session threads endpoint (`GET
/api/v2/aibridge/sessions/{session_id}`), so the session page can show
cost alongside token usage. Backend only, the UI changes for AIGOV-600
will follow separately.

## Changes

- Add `cost_micros` and `has_unpriced_usage` to
`AIBridgeSessionThreadsTokenUsage`, populated for the session summary,
each thread, and each agentic action.
- Compute the session summary cost in `ListAIBridgeSessions`, next to
the existing session token sums, so tokens and cost come from the same
rows.
- Add `priced_model`, the model and per-token prices used for pricing,
to `AIBridgeThread` for the root interception (the prompt) and to each
`AIBridgeAgenticAction` for that interception, since requests in a
thread may be priced as different models. Nil when not priced.
- Update frontend mocks for the new required fields.

Related to AIGOV-600

> [!NOTE]
> Generated by Coder Agents on behalf of @ssncferreira.
2026-10-26 10:01:50 +00:00
Sas Swart 946c2f0e04 feat: resolve orgs from workspace session tokens in AI Gateway requests (#30902)
Add an AI Gateway org resolver at the coderd IsAuthorized boundary,
detect workspace session tokens, load the workspace org, validate the
user is still a member of that org, reject missing workspaces or invalid
membership, and return the resolved org in IsAuthorizedResponse while
leaving it unused by downstream policy. Include tests for successful
workspace-token resolution, missing workspace rejection, invalid org
membership rejection, and compatibility with the existing ServeHTTP
flow.

Fixes AIGOV-714
2026-10-26 11:40:51 +02:00
Nick Vigilante 9edb2e388d feat: enable Coder.USSpelling Vale rule (#30790)
Depends on #30782.

Adds `Coder.USSpelling` at `warning` severity.
The rule flags an `-ise` family (`organise`, `initialise`,
`authorisation`) and a list of British spellings (`behaviour`,
`labelled`, `cancelled`, `analogue`, `whilst`, and others) and suggests
the US form.
It covers `docs/**` except `docs/.style/style-guide/**` and the
generated `docs/reference/{api,cli}/*.md` pages.

Baseline: 13 findings before cleanup, 0 after.
The first commit fixes 11 of them: 10 in hand-written pages and 1 in the
Prometheus metrics table, which comes from metric help text.
Two findings sit in generated reference pages, which `.vale.ini`
excludes for this rule: the CLI alias `favourite` and `amongst` in a Go
comment from a vendored dependency (the DERP map schema).
The metric help text comes from `querymetrics.go`, so the fix lands
there and in `generated_metrics` and `prometheus.md`.

Known false positive: Vale doesn't skip bold text, so a UI label in bold
that keeps British spelling is flagged.
The rule comment and the guide footer say to use a code span for such a
label.

Linear: DOCS-1280

<details>
<summary>Decision log</summary>

- **Rule type**: `substitution`, so each message names the US form.
The `-ise` family can't name a specific replacement, so it suggests "the
-ize form".
- **`cancelled`**: the guide's "Use US spelling" section lists
`canceled` and `canceling`.
The rule flags the double-l verb forms and leaves `cancellation` alone,
which Merriam-Webster spells with one `l`.
- **Exceptions**: Vale applies an exception to the matched word, so each
is anchored with `^` and `$` and case-insensitive.
The list covers `-wise` words and stems for `advise`, `advertise`,
`exercise`, `promise`, `premise`, `compromise`, `surprise`, `comprise`,
`concise`, `precise`, `revise`, `supervise`, `televise`, `devise`,
`improvise`, `franchise`, `merchandise`, `disguise`, `despise`,
`paradise`, `cruise`, `noise`, `raise`, `rise`, `wise`, and others, each
with the endings `e`, `es`, `ed`, `ing`, `er`, and `ers`.
  The `mise` tool is excepted by name.
- **Extra British spellings** beyond the guide table: other doubled-`l`
verbs (`signalled`, `levelled`), `metre`, `litre`, `fibre`, `theatre`,
`calibre`, `offence`, `pretence`, `analyse`, `judgement`,
`acknowledgement`, `ageing`, `artefact`, `sceptic`, `aluminium`,
`practise`, `enrol`, `fulfil`, `instil`, `skilful`.
  None appears in the corpus today.
- **Generated pages**: excluded for `docs/reference/{api,cli}/*.md`, as
`Coder.UnitSpacing` does.
Changing the `favourite` alias would change CLI behavior, and the
`amongst` comment isn't ours to edit.
- **Prometheus doc**: edited by hand with matching table padding because
`go run ./scripts/metricsdocgen` already rewrites the whole table on the
base branch, independent of this change.

Test cases, run with `vale --config` against fixtures that are not
committed, in LF and CRLF:

- Positive: `organise`, `initialise`, `authorise`, `customisation`,
`authorisation`, `organisations`, `Organised`, `initialising`,
`recogniser`, `customisable`, every explicit list entry, and capitalized
forms.
- Negative: `advise`, `exercise`, `otherwise`, `Enterprise`, `premise`,
`compromise`, `promise`, `surprise`, `precise`, `concise`, `expertise`,
`revising`, `supervising`, `improvising`, `raise`, `noise`, `cruise`,
`franchise`, `merchandise`, `disguise`, `despise`, `paradise`,
`appraise`, `sunrise`, `arise`, `liaise`, `clockwise`, `cancellation`,
`canceled`, `mise`.
- Skipped: words in code spans and fenced code blocks, such as
`cancelled`.
- LF and CRLF results are identical.

Checks: `vale --output=line docs/` reports 0 `Coder.USSpelling`
findings, `make lint/style-claims` passes, and the pre-commit hooks pass
on both commits.

</details>

*Generated by [Coder Agents](https://coder.com/docs/ai-coder/agents) on
behalf of @nickvigilante.*
2026-10-23 22:18:30 -07:00
c4d76a4009 feat: add skill action and automatic-use settings (#30882)
Adds two per-skill settings that the project skills UI will expose next:

- `action` marks a project skill to show as an action button above
project chat composers. A `skills_action_project_only` check keeps it to
project skills; the personal and organization routes reject `action`
with 400.
- `model_invocable` (default true, every stored source) controls
automatic use. When it's false the skill is left out of
`<available-skills>` and `read_skill` reports it as not found, but a
structured skill invocation (a chip from the `/` menu or an action)
still loads it, as long as the skill is enabled and the chat owner can
read it. Workspace skills always count as model-invocable.

Migration 000625 adds both columns with defaults, so existing skills
keep today's behavior. The list, get, insert and update queries,
`codersdk.SkillMetadata` and the create/update requests carry both
fields, and the audit table tracks them.
`coderd/x/chatd/ARCHITECTURE.md` gets TODO notes for the author.

**NOTE**: the migration adds two columns with constant defaults and a
CHECK constraint to `skills`.

Stack (merge top to bottom):
- #30826
- #30827
- #30838
- #30840
- #30854
- #30855
- #30856
- #30861
- #30880
- #30881
- #30882 👈
- #30884
- #31122
- #31126

---------

Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Nick Vigilante <nickvigilante@users.noreply.github.com>
2026-10-23 19:32:08 +07:00
Ethan 9b8d880d63 fix: exempt service accounts from user dormancy (#30824)
Closes CODAGT-1344

The Slack bot for Coder Agents (CODAGT-1098) runs inside coderd and acts
as a service account that an admin picks. It creates chats and sends
messages as that account through Go calls, not HTTP requests. Two
dormancy rules break it:

1. **New service accounts start dormant.** The CLI and the web "Create
user" page send no status, and `InsertUser` falls back to `dormant`. So
the bot's account is dormant from the start.
2. **Accounts go dormant after 90 days without a "last seen" update.**
Only the API key middleware updates `last_seen_at`, and only on HTTP
requests. The bot never sends any, so its account goes dormant on day
90, even if people use the bot every day.

A chat owner must be active, so once the bot's account is dormant, every
Slack mention fails. Any other in-process integration that acts as a
service account has the same problem.

Dormancy does nothing useful for service accounts:

- It exists to free license seats, and service accounts never take a
license seat (#24401, #27279).
- It isn't a lock. Any API request activates a dormant account again.
- System users already skip the dormancy job for the same reason.

So this PR:

- creates service accounts as active when the request doesn't set a
status,
- rejects an explicit dormant status for a service account with a 400,
- skips service accounts in the dormancy job, the same way it already
skips system users.

Side effects: a service account that nobody uses for 90 days now stays
active. That has no effect on license seats. If the account has ever
sent traffic through AI Gateway, it also keeps its AI Governance seat
until an admin suspends or deletes it. That matters less than it sounds:
AI Governance is now included with Premium rather than sold per seat
(#27545), so the seat only counts on a license that still has an AI
Governance seat limit, and there it only feeds the seat-limit warning.
The dormancy job also no longer writes "marked dormant" audit entries
for service accounts.

I've left existing dormant service accounts alone rather than adding a
migration. An admin can activate one, and any API request from it
activates it too.

---

_This PR description was written by Coder Agents._
2026-10-23 12:53:40 +11:00
Sas Swart 837c2e1af1 feat(coderd): price token usage with organization prices (#30719)
Prices token usage by the most specific match: the organization's price
for the configured provider, then a site-wide custom price, then the
price book. The requested model is still preferred over the model the
provider reports.

- `GetAIModelPriceForProvider` replaces
`GetAIModelPriceByProviderModel`, which is removed. It is one `UNION
ALL` of primary-key lookups, ordered by requested model first, then by
specificity. dbauthz checks `ResourceAiModelPrice`, scoped to the org
when one is given.
- `coderd/aibridge/prices` tests read seeded prices through
`GetAIModelPrices`.

Interceptions don't carry an organization yet, so
`resolveTokenUsageCost` passes none and only site-wide prices apply at
runtime for now (AIGOV-716). Org matching is covered by database tests
with explicit org IDs. For the same reason, the unpriced-models report
still counts only site-wide prices.

**Used later:** prices set through the API in #30720 take effect here
once interceptions carry an organization.

---

*This PR was generated by Coder Agents on behalf of @SasSwart.*
2026-10-22 15:13:24 +00:00
Sas Swart da78c88100 feat(coderd): add queries for organization AI model prices (#30718)
Adds the store queries for organization AI model prices in
`ai_model_configs`, with authorization. They cover basic CRUD on one
price, keyed by `(organization_id, provider_id, model)`.

- `GetOrganizationAIProviderModelPrice` reads one price.
- `UpsertOrganizationAIProviderModelPrice` sets one price, with typed
params. It rewrites an existing row only when a price differs, so
`updated_at` records the last change.
- `DeleteOrganizationAIProviderModelPrice` soft deletes one price. Reads
and upserts only match live rows, so setting a price after a delete
inserts a new live row.

All three authorize against `ResourceAiModelPrice.InOrg(org)`. The new
`AiModelPriceInOrg` case in `TestRolePermissions` pins that only owners
and org admins pass. A TODO in dbauthz notes that this should become a
resource for AI Governance model config.

Each query has its own test. The upsert tests cover that identical
prices leave `updated_at` unchanged, that changed prices advance it, and
that an empty model is rejected.

**Used later:** #30719 seeds its lookup tests with the upsert. #30720
serves all three through the org price API.

---

*This PR was generated by Coder Agents on behalf of @SasSwart.*
2026-10-22 13:58:42 +00:00
Michael Suchacz 98104d10fe chore(coderd/database): drop the chat debug stale-sweep indexes (#31009)
Drops `idx_chat_debug_runs_stale` and `idx_chat_debug_steps_stale`, the
partial `updated_at` indexes that migration 000468 added for the
`FinalizeStaleChatDebugRows` query. #30995 deleted that query, so
nothing reads these indexes, but every debug run and step write still
updates them.

**Why it is safe:** no remaining query can use them. A partial index on
`updated_at WHERE finished_at IS NULL` only serves queries that filter
on `finished_at IS NULL`, and the one query that still does, the
`locked_run` CTE in `InsertChatDebugStep`, finds the run by primary key.
`DeleteOldChatDebugRuns` filters on `updated_at` alone and uses
`idx_chat_debug_runs_updated_at` from migration 000487, which stays. The
down migration recreates both indexes as 000468 defined them.

**Review focus:** the migration number. 000632 follows `main`'s newest
migration, 000631 from #31059. If `main` gains another migration first,
this one moves to the next free number.

Validation:
- `make gen` changes only `dump.sql`, which loses the two `CREATE INDEX`
lines. The renumber to 000632 leaves it unchanged.
- `go test -count=1 ./coderd/database/migrations/` passes on this
branch, where `main`'s 000631 and this 000632 both apply. The run
includes `TestMigrate/UpDownUp` (every migration up, down and up again)
and `TestMigrationChain`.
- The pre-commit hook passed.

## Stack

This stack deletes the stale debug-run sweep, then simplifies the
chatdebug debug-run, stream wrapper and step numbering code, drops the
indexes the sweep used, and last deletes a runner debug field that lost
its only reader with the heartbeats. The debug-run and stream wrapper
refactors each sit on a test-only layer that pins the behavior they
keep.

1. #30988 Delete the chat debug run and step heartbeats
2. #30989 Delete `FinalizeStale` and the stale threshold
3. #30995 Delete the `FinalizeStaleChatDebugRows` query
4. #30933 Pin the debug runs that turns, titles and status labels record
5. #30934 Create and finalize debug runs through one helper
6. #30980 Table the stream wrapper tests and cover their status rules
7. #30981 Share one finalize-once lifecycle between the stream wrappers
8. #31007 Number chat debug steps in the database
9. #31008 Delete the process-global step counters
10. This PR: Drop the chat debug stale-sweep indexes
11. #31010 Drop the unused `runnerCtx` from the runner debug turn

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-22 07:42:22 +00:00
Hugo Dutka c61d842b9c feat: show available ACP harnesses in chat context (#31059)
Expose discovered ACP harnesses in the chat context resource inventory
when the `chat-acp-subagents` experiment is enabled, including
configuration paths and resource status. Persist ACP harness resources
from agent context snapshots and update the SDK and generated API types.

Generated with Codex, carefully reviewed by Hugo.
2026-10-21 14:18:12 +00:00
Michael Suchacz eea839ccc4 refactor(coderd): number chat debug steps in the database (#31007)
`InsertChatDebugStep` now numbers a step after the run's highest step
when the caller passes zero, so chatdebug stops numbering steps in
process memory. `CreateStepParams` loses `StepNumber`, `beginStep` no
longer calls `nextStepNumber` or `syncStepCounter`, and `CreateStep`
replaces its collision retry, which listed the run's steps and bumped
the number, with a per-run lock. #31008 deletes the counters, which no
production code reads after this PR.

**Why it is safe:** Postgres takes the insert's snapshot before the
statement waits on the run row, so creates that waited on the same run
read the same highest step. `CreateStep` therefore takes
`pg_advisory_xact_lock(LockIDChatDebugRunSteps(run))` in its own
statement and inserts in the same transaction, so each insert starts
after the previous holder committed. The new
`TestService_CreateStep_NumbersConcurrentSteps` holds the run row,
starts 20 creates, waits until all of them wait on a lock, then commits
and checks the numbers 1 to 20. It passes 10 of 10 runs, and fails 10 of
10 with a 10-attempt collision retry in place of the lock.
`TestService_CreateStep_RetriesDuplicateStepNumbers` becomes
`TestService_CreateStep_NumbersStepsPerRun`, and
`TestService_CreateStep_ListRetryErrorWins` goes with the list call it
covered. Mutants that fail the chatdebug tests: no lock, the lock
outside the transaction, the lock keyed by chat, a constant number,
storing the zero, numbering across runs, a nonzero number from
`CreateStep`, and an unclassified missing run. One mutant survives and
is equivalent: running the insert on `s.db` instead of `tx` inside the
same `InTx` callback, because the insert still commits before the
transaction releases the lock. Running the insert after the transaction
commits fails the new test with a duplicate key. A copy of `service.go`
with garbage in it fails to build.

**Review focus:** `@step_number` stays as an override. Production passes
zero, and the fixtures in `querier_test.go`, `dbauthz_test.go`,
`dbpurge_test.go` and `exp_chats_test.go` keep their explicit numbers.
No test depends on the override (a mutant that ignores it survives),
because every fixture numbers its steps in insert order. Dropping the
parameter would touch about 20 fixtures outside chatdebug.

The PR removes more than it adds: +179/-233 hand-written lines. The
generated `querier.go` and `queries.sql.go` change 15 more.

Validation:
- `make gen/db` changes only `querier.go` and `queries.sql.go`.
- `go test -race -count=1 -p 4 -parallel 4
./coderd/x/chatd/chatdebug/...` passes, and `go test -count=10 -run
TestService_CreateStep ./coderd/x/chatd/chatdebug/` passes all 6 tests
10 of 10 times.
- `go test -count=1 -run ChatDebug ./coderd/database/
./coderd/database/dbpurge/` and `go test -count=1 -run Debug
./coderd/x/chatd/` (19 tests) pass, and the 25 chatd debug-run tests
pass under `-race`.
- `golangci-lint run --fix=false ./coderd/x/chatd/chatdebug/...
./coderd/database/` reports 0 issues, and the pre-commit hook passed.

## Stack

This stack deletes the stale debug-run sweep, then simplifies the
chatdebug debug-run, stream wrapper and step numbering code, drops the
indexes the sweep used, and last deletes a runner debug field that lost
its only reader with the heartbeats. The debug-run and stream wrapper
refactors each sit on a test-only layer that pins the behavior they
keep.

1. #30988 Delete the chat debug run and step heartbeats
2. #30989 Delete `FinalizeStale` and the stale threshold
3. #30995 Delete the `FinalizeStaleChatDebugRows` query
4. #30933 Pin the debug runs that turns, titles and status labels record
5. #30934 Create and finalize debug runs through one helper
6. #30980 Table the stream wrapper tests and cover their status rules
7. #30981 Share one finalize-once lifecycle between the stream wrappers
8. This PR: Number chat debug steps in the database
9. #31008 Delete the process-global step counters
10. #31009 Drop the chat debug stale-sweep indexes
11. #31010 Drop the unused `runnerCtx` from the runner debug turn

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-21 08:23:28 +00:00
Sas Swart d14631df1a feat(coderd/database): add organization AI model config table (#30717)
Adds `ai_model_configs`: an organization's AI Governance configuration
for one model on one configured provider. Token prices are its first
settings. Migration `000630` is additive and needs no backfill.

- Rows are soft deleted with a `deleted` flag. A surrogate `id` is the
primary key, and a partial unique index allows one live row per
`(organization_id, provider_id, model)`. Cascades on delete of the
organization or provider. The organization is a column, as in other
organization-scoped tables.
- A check constraint rejects an empty model.
- Pricing per provider ID is what lets `openai-compat` gateways be
priced, and lets providers of the same type have different prices.
- Site-wide prices per provider type stay in the existing
`ai_model_prices` table, unchanged.

Recording which price was used on each token usage record is deferred
until its design settles.

**Used later:** #30718 adds the queries that read and write the table,
and #30719 looks prices up from it. The rest of the diff is generated.

---

*This PR was generated by Coder Agents on behalf of @SasSwart.*
2026-10-21 06:17:12 +00:00
Michael Suchacz 8b6f209e8f refactor: delete the FinalizeStaleChatDebugRows query (#30995)
Deletes the `FinalizeStaleChatDebugRows` query. #30989 deleted
`chatdebug.Service.FinalizeStale`, its only caller. This is the last of
three PRs that delete the stale debug-run sweep.

- Deletes the query from `chatdebug.sql` and regenerates the querier,
dbmock and dbmetrics code. `make gen/db` leaves the dbmetrics wrapper of
a deleted query behind, so I removed it by hand.
- Deletes the dbauthz wrapper and its `TestMethodTestSuite` case, and
`TestFinalizeStaleChatDebugRows` from `querier_test.go`.
- The `codersdk.ChatDebugStatus` comment no longer names the query.

**Why it is safe:** nothing calls the query after #30989. #30988
explains why no reader of the debug `updated_at` columns needs the sweep
and why dbpurge cannot delete a live run.

**Review focus:** the partial indexes `idx_chat_debug_runs_stale` and
`idx_chat_debug_steps_stale` from migration 000468 served only this
query. They stay in this PR; #31009, at the top of this stack, drops
them in a migration.

Validation: `make gen/db` and `go build ./...`. `go test -count=1`
passes for `./coderd/database/dbauthz/ -run TestMethodTestSuite`,
`./coderd/database/ -run ChatDebug`, `./codersdk/` and
`./coderd/database/dbpurge/`. The pre-commit hook passed.

## Stack

This stack deletes the stale debug-run sweep, then simplifies the
chatdebug debug-run, stream wrapper and step numbering code, drops the
indexes the sweep used, and last deletes a runner debug field that lost
its only reader with the heartbeats. The debug-run and stream wrapper
refactors each sit on a test-only layer that pins the behavior they
keep.

1. #30988 Delete the chat debug run and step heartbeats
2. #30989 Delete `FinalizeStale` and the stale threshold
3. This PR: Delete the `FinalizeStaleChatDebugRows` query
4. #30933 Pin the debug runs that turns, titles and status labels record
5. #30934 Create and finalize debug runs through one helper
6. #30980 Table the stream wrapper tests and cover their status rules
7. #30981 Share one finalize-once lifecycle between the stream wrappers
8. #31007 Number chat debug steps in the database
9. #31008 Delete the process-global step counters
10. #31009 Drop the chat debug stale-sweep indexes
11. #31010 Drop the unused `runnerCtx` from the runner debug turn

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-20 21:10:40 +00:00
Michael Suchacz 367358862b refactor: delete chatdebug FinalizeStale and the stale threshold (#30989)
Deletes `chatdebug.Service.FinalizeStale` and everything that existed
only for it. Nothing in production calls `FinalizeStale` since #26270,
and #30988 removed the heartbeats that kept live runs out of its reach.
This is the second of three PRs that delete the stale debug-run sweep.

- `chatdebug/service.go`: deletes `FinalizeStale`, `SetStaleAfter`,
`DefaultStaleThreshold`, the `staleAfterNanos` field and `WithClock`.
The service keeps its real clock for the timestamps it writes.
- `chatd.go`: the debug service factory no longer calls
`SetStaleAfter(inFlightChatStaleAfter * 3)`. `inFlightChatStaleAfter`
stays, because the chat worker uses it.
- `chatdebug/types.go`: deletes `EventKindFinalize` and
`BroadcastPubsubChannel`. `FinalizeStale` was the only publisher without
a chat ID, so `PubsubChannel` loses its `uuid.Nil` branch. Nothing in
production subscribes to `chat_debug:*` channels.
- Deletes the four `FinalizeStale` service tests, and
`TestFinalizeStaleChatDebugRows_TerminalStatusAlignment` together with
`codersdk.ChatDebugTerminalStatuses`, which only that test used. The
test compared that list with the `NOT IN` lists in
`FinalizeStaleChatDebugRows` and nothing else: no other debug query or
constraint lists the statuses. The next PR deletes that query.
- Comments that named the stale sweep now say what happens instead:
`debugCleanupClockSkew`, `errRunFinalized`, `CreateStep`, `UpdateStep`,
the `InsertChatDebugStep` and `UpdateChatDebugStep` queries, the edit
cleanup in `chat_mutations.go`, the two notes on removed tests in
`chatd_internal_test.go`, and four comments in `chatd_test.go`.

**Why it is safe:** `FinalizeStale` and `SetStaleAfter` had no
production caller other than the factory line this PR removes, and
nothing calls `WithClock` after #30988. `make gen/db` changes only the
regenerated query comments. #30988 explains why no reader of
`updated_at` needs the sweep and why dbpurge cannot delete a live run.

**Review focus:** the `debugCleanupClockSkew` comment in
`chatd_debug.go`. The value stays `30 * time.Second`. Runs that start
inside that buffer survive the fast cleanup; the old comment said the
stale sweep would then finalize them, but dbpurge deletes them once they
age past the debug retention window.

Validation: `make gen/db`, `go build ./...`, and `go vet` on the touched
packages. `go test -count=1` passes for
`./coderd/x/chatd/chatdebug/...`, `./coderd/x/chatd/ -run
'Debug|EditMessage|ArchiveChat'`, `./coderd/database/ -run
'ChatDebug|TestIsAuthorizedQuery|TestInsertChatMessagesOrderContract|ConnectionLogs'`
and `./codersdk/`. The pre-commit hook passed.

## Stack

This stack deletes the stale debug-run sweep, then simplifies the
chatdebug debug-run, stream wrapper and step numbering code, drops the
indexes the sweep used, and last deletes a runner debug field that lost
its only reader with the heartbeats. The debug-run and stream wrapper
refactors each sit on a test-only layer that pins the behavior they
keep.

1. #30988 Delete the chat debug run and step heartbeats
2. This PR: Delete `FinalizeStale` and the stale threshold
3. #30995 Delete the `FinalizeStaleChatDebugRows` query
4. #30933 Pin the debug runs that turns, titles and status labels record
5. #30934 Create and finalize debug runs through one helper
6. #30980 Table the stream wrapper tests and cover their status rules
7. #30981 Share one finalize-once lifecycle between the stream wrappers
8. #31007 Number chat debug steps in the database
9. #31008 Delete the process-global step counters
10. #31009 Drop the chat debug stale-sweep indexes
11. #31010 Drop the unused `runnerCtx` from the runner debug turn

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-20 20:54:18 +00:00
Michael Suchacz 2397f3953a refactor(coderd): delete the chat debug run and step heartbeats (#30988)
Deletes the chat debug heartbeats. Model calls and streams started a
goroutine that touched the step and its run on a ticker, and runner
debug turns started a second one that touched the run. Both existed so
that `FinalizeStale` would not mark a live run as interrupted, but
nothing in production calls `FinalizeStale` since #26270. Mike chose to
delete the stale sweep as dead code. This PR is the first of three that
do it.

- `chatdebug/model.go`: `Generate`, `GenerateObject` and both stream
wrappers no longer start `launchHeartbeat`.
- `active_turn_debug.go`: the runner debug turn no longer starts
`LaunchRunHeartbeat`.
- `chatdebug/service.go`: deletes `TouchStep`, `TouchRun`,
`LaunchRunHeartbeat`, `heartbeatInterval`, and the channel that
`SetStaleAfter` closed to wake the heartbeats. The `WithClock` doc no
longer names the heartbeat tickers.
- Deletes the `TouchChatDebugStepAndRun` and
`TouchChatDebugRunUpdatedAt` queries with their dbauthz wrappers and
tests, and `TestLaunchHeartbeat`. `TestFinalizeStaleChatDebugRows` now
backdates its run with `UpdateChatDebugRun` instead of the deleted
test-only query.

**Why it is safe:** the heartbeats wrote only `updated_at` on
`chat_debug_runs` and `chat_debug_steps`. These read it:

| Reader | Why it does not need the heartbeat |
|---|---|
| `FinalizeStaleChatDebugRows` and the partial indexes
`idx_chat_debug_runs_stale` and `idx_chat_debug_steps_stale` | No
production caller. The next two PRs delete the query, and #31009 drops
the indexes in a migration. |
| `DeleteOldChatDebugRuns` (dbpurge) | See the next paragraph. |
| `DebugPanel.tsx` run sort | Sorts by `started_at \|\| updated_at`, and
`started_at` is `NOT NULL`, so it never falls back to `updated_at`. |
| Debug run and step API responses | Return `updated_at` as data.
Nothing uses it to decide whether a run is alive. |

dbpurge deletes runs whose `updated_at` is older than the debug
retention window, whatever their status, and their steps cascade. The
window is `agents_chat_debug_retention_days`: 30 days by default, and 0
turns the purge off, so dbpurge never deletes a run that was written in
the last day. A run's `updated_at` still moves when the run is created,
when each step is inserted (the `InsertChatDebugStep` CTE), and on every
run update and finalize. So dbpurge could only delete a live run if one
model call plus the tool calls after it ran for more than a day. The
runner finalizes its debug run when the runner context ends, a
`requires_action` wait times out after 5 minutes (the case `TouchRun`
covered), and chatloop aborts a stream that is silent for
`--chat-stream-silence-timeout`, 10 minutes by default (0 turns that
guard off).

Mike accepted the cost: a run or step orphaned by a crashed replica
stays `in_progress` until dbpurge deletes it. That is already true on
`main`, because nothing runs the sweep.

This does not touch the chat worker's `chat_heartbeats` loop that
ARCHITECTURE.md describes, and no ARCHITECTURE.md sentence about debug
runs becomes false.

**Review focus:** nothing uses `WithClock` after this PR, but deleting
it here would take this PR past 500 changed lines, so it goes with
`FinalizeStale` in the next PR.

Validation: `make gen/db`, `go build ./coderd/...`, and `golangci-lint
run` on the touched packages (0 issues). `go test -count=1` passes for
`./coderd/x/chatd/chatdebug/...`, `./coderd/database/dbauthz/ -run
TestMethodTestSuite`, `./coderd/database/ -run ChatDebug`,
`./coderd/x/chatd/ -run Debug` and `./coderd/database/dbpurge/`. The
pre-commit hook passed.

## Stack

This stack deletes the stale debug-run sweep, then simplifies the
chatdebug debug-run, stream wrapper and step numbering code, drops the
indexes the sweep used, and last deletes a runner debug field that lost
its only reader with the heartbeats. The debug-run and stream wrapper
refactors each sit on a test-only layer that pins the behavior they
keep.

1. This PR: Delete the chat debug run and step heartbeats
2. #30989 Delete `FinalizeStale` and the stale threshold
3. #30995 Delete the `FinalizeStaleChatDebugRows` query
4. #30933 Pin the debug runs that turns, titles and status labels record
5. #30934 Create and finalize debug runs through one helper
6. #30980 Table the stream wrapper tests and cover their status rules
7. #30981 Share one finalize-once lifecycle between the stream wrappers
8. #31007 Number chat debug steps in the database
9. #31008 Delete the process-global step counters
10. #31009 Drop the chat debug stale-sweep indexes
11. #31010 Drop the unused `runnerCtx` from the runner debug turn

> Xum, an AI agent, opened this PR for @ibetitsmike and wrote this
description.
2026-10-20 20:32:25 +00:00
Faur Ioan-AurelandClaude Opus 5.5 1a65385a4b fix: log workspace shutdown reasons for SSH connections (#30515)
Workspace stops and deletes closed active SSH connections with generic
shutdown or process-exit reasons. This obscured the workspace action
that ended a connection and made related logs harder to correlate.

The agent’s SSH tracking and logging now identify connections closed by
the SSH server as server_shutdown, while preserving earlier client or
network disconnects. The agent includes the client session ID when
available for ordinary sessions and connections used only for port
forwarding.

Coderd now enriches shutdown reasons when it processes connection
reports. It translates server_shutdown into workspace_stopped when the
latest build is a running or successful stop or delete requested before
the connection closed. Coderd persists the enriched reason and logs the
build reason, transition, and build ID alongside the client session ID.
Unavailable or unrelated build state retains the generic shutdown
reason.

The CLI now logs shutdown context once per stop or delete build through
its existing workspace watcher. This includes the client session ID when
available and preserves the existing graceful SSH shutdown behavior.

Regarding the PTY connections, when the agent shut down, the
reconnecting PTY server closed each PTY
and its attached connections, and the connection handler then returned
without an error. The server therefore logged and reported those
connections as graceful closes. Its fallback for a forced shutdown
reported a free-form reason that coderd did not recognize, so coderd
never refined either case into a workspace shutdown.

We change the reconnecting PTY server so that it reports any connection
that ends after the agent begins shutting down as a server shutdown
initiated by the agent, and its log keeps the client session ID. The
forced shutdown fallback now reports the same standard reason, which
lets coderd translate it into a workspace stop or delete in the
connection log, as it already does for SSH. A client that closes its
connection before the shutdown still keeps its graceful reason.

Regression tests cover shutdown attribution, fallback behavior, CLI
logging, and PostgreSQL persistence with and without a client session
ID. The test logging helper also synchronizes with cleanup to prevent
logging after a test completes

- resolves
https://linear.app/codercom/issue/CLIENT-665/close-workspace-tunnel-with-a-reason

<!--

If you have used AI to produce some or all of this PR, please ensure you
have read our [AI Contribution
guidelines](https://github.com/coder/coder/blob/main/contributing/AI_CONTRIBUTING.md)
before submitting.

-->

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-19 23:06:41 +03:00
Michael Suchacz 200fa7dea7 perf(coderd): bound the initial chat stream snapshot to the client cursor (#29761)
Opening a chat on the Agents page loads its transcript over REST and
then connects the chat stream with `after_id` set to the newest loaded
message. The first sync of the stream ignores that cursor when it
queries the database: it fetches every row of the chat on each
connection, and then sends only the rows above `after_id`.

The first sync now resolves `after_id` to that message's revision and
fetches only rows changed at or after it. The history reset rule is the
one on `main` (#29808): the stream resets only when message `after_id`
is deleted or missing. A deleted cursor still produces one
`history_reset` with the surviving history, because the bounded query
includes the cursor row. An unknown ID keeps the full query. A message
ID from another chat, live or deleted, is ignored, so the client
receives the full history. Live updates after the first sync are
unchanged.

## Performance

Measured on dogfood's database, which runs `main`, with `EXPLAIN
(ANALYZE, BUFFERS, SERIALIZE BINARY)` of the first-sync query. The chats
are the 554 chats updated in the past day, picked by row count. `main`
starts at revision 0. This PR starts at the revision of the newest live
message, which is the `after_id` the web client sends.

| Chat (rows) | `main` output | `main` first / repeat run | This PR
output | This PR time |
|---|---|---|---|---|
| p50 (41) | 189 kB | 21.5 ms / 0.6 ms | 1 kB | 0.1 ms |
| p90 (351) | 2.0 MB | 163 ms / 7.9 ms | 10 kB | 0.5 ms |
| p99 (2,200) | 7.1 MB | 425 ms / 25 ms | 14 kB | 3.0 ms |
| Largest (14,746) | 64.9 MB | 4.8 s / 193 ms | 6 kB | 17.6 ms |

- Across the 554 chats, the bounded query returns 1 row up to p99 and 5
rows at most.
- Both queries visit the same heap pages, because no index covers
`revision`. The saving is the TOAST reads, serialization, transfer, and
decoding of rows that `main` then discards. Most of `main`'s time is
serialization: 153 ms of 193 ms for the largest chat.
- The cursor lookup is one primary-key read of about 0.4 ms, so a median
chat gains almost no latency. The gain grows with chat size.
- Deleted rows are about 0.1% of rows, so they do not change the cost.

## Not changed

A tab whose stream was down can miss an edit, then send a message and
store the response, so its newest ID is above the edit. On reconnect,
message `after_id` is live and the stream sends no reset, so the tab
keeps the deleted messages. `main` has done the same since #29808, and
this PR does not change it (mafredri's reproduction gives `expected [5 6
7 8], actual [3 4 7 8]` on both). An ID cannot show which edits the
client applied. #29662 adds a client-reported version for that.

<details>
<summary>Delivery record</summary>

- Rebased on `main` at `6ebefb83254`, keeping the #29808 reset rule and
#30266 read-only snapshot reads. The cursor-bound mock test is part of
`TestStreamLoopInitialSyncTombstones`.
- Assumption checked on dogfood: no live message has a higher revision
than a later message. Messages are only updated to soft-delete them or
to set search columns, which the revision trigger ignores. 2,861 rows in
29 recently active chats break it, all created between 2026-06-26 and
2026-07-16 and none since. A client cursor cannot sit before them,
because the client sends its newest message.
- Local: `go test ./coderd/x/chatd/...`, `go test ./coderd -run
'^TestStreamChat$'`, `go test ./coderd/database/dbauthz -run
'^TestMethodTestSuite$'`. Mutants of the cursor bound (no `- 1`, no
bound, cursor from another chat kept) each fail `TestStreamLoop*`.
- Remote UAT before the rebase covered deleted, foreign, and unknown
cursors, edits, grouped commits, and reconnects. On a synthetic
5,000-message chat, dial to first snapshot went from 129.6 ms to 15.0
ms.
- Codex: [round
1](https://github.com/coder/coder/pull/29761#pullrequestreview-5284702021)
and [round
2](https://github.com/coder/coder/pull/29761#pullrequestreview-5284855535)
found cross-chat cursor bugs, fixed. Round 3 and the round on
`7bed7855e41` found nothing.

</details>

> Xum opened this PR on behalf of @ibetitsmike.
2026-10-19 19:15:56 +02:00
Ehab Younes f4cd0deccb feat!: report template insights session usage per app (#29796)
Report template insights session usage per app instead of five fixed
family rows, with registry names and icons.

- Add per-app rows for reported session apps (VS Code, Cursor, GoLand,
  Zed, etc.); unused apps are omitted
- Replace `type` (`builtin`/`app`) with `method` (`ssh`,
  `reconnecting_pty`, `port_forwarding`, `workspace_app`); session
  methods are derived from the app name until #31012 stores them
- Add `usage_total_seconds` to count each user's active minutes once,
  and use it as the UI bar denominator so overlapping apps can exceed
  100%
- Merge template apps sharing a slug into one row across templates,
  using the newest display name and icon
- Rename Web Terminal slug from `reconnecting-pty` to
  `reconnecting_pty`
- Cap the response at the 64 busiest unregistered apps and fold the
  rest into "Other apps", flagged with `seconds_is_upper_bound`
  ("Up to" in the UI)
- Add `unknown` and `overflow` accounting rows with explanatory
  tooltips, and a fallback icon for apps without one
- Remove the unused `codersdk.UnionByFamily` helper from the Go SDK

BREAKING CHANGE: `report.apps_usage` in `GET /api/v2/insights/templates`
no longer returns the fixed five rows, family slugs, or previous display
names. Consumers must update.

Part of CLIENT-642
Prometheus metrics handled separately in #29906
2026-10-19 17:59:18 +03:00
Michael Suchacz 0968f97a43 feat: move a chat family between projects in chatd (#30893)
chatd can now move a root chat, together with its subagent chats, into a
project, to another project, or out of its project. Nothing calls this
yet. #30894 adds the API.

**Stack** (moving chats between projects, bottom first):
1. **This PR**: the move in chatd
2. #30894: `PATCH /api/v2/chats/{chat}` takes `project_id`
3. #30895: **Move to project** in the chat menu
4. #30795: drag a chat onto a project folder in the sidebar

**Why:** today a chat keeps the project it was created with. A move must
follow the family rules of archiving, because a step that runs during a
move would still have the memory tools of the old project.

- `chatstate.Tx.SetProject` runs only from W, E0 and E1, like
`SetArchived`. `SetFamilyProject()` locks the root row before its
subagents, in the same order as archiving, so a busy family gets
`ErrTransitionNotAllowed` and nothing changes.
- A move to the project the chat is already in changes nothing, even
while the chat runs. The check reads the locked root row, so it waits
for a move that is still in flight.
- Only the root row stores `project_id`, and subagents use the project
of the root. After the commit, chatd publishes a new `project_change`
watch event for each chat of the family. The move keeps `updated_at`, so
lists keep their order.
- A root chat outside a project now has an empty memory index. When a
chat leaves its project, or its project is deleted, its next turn start
commits one memory index update that removes the memories the model saw
there.

**What to scrutinize:** the lock order and the same-project check under
the lock in `SetFamilyProject()` (`chatstate/family.go`), the turn-start
condition in `memoryIndexMessage()` (`memory.go`), and whether the
`ARCHITECTURE.md` text matches them.

**Generated** (`make gen`): `queries.sql.go`, `querier.go`, `dbmock` and
`dbmetrics` come from `queries/chats.sql`. `coderd/apidoc`,
`docs/reference/api/schemas.md` and `typesGenerated.ts` come from the
new watch event kind, which is also why two `chats.test.ts` fixtures
gain `project_change: false`.

**Architecture doc:** `coderd/x/chatd/ARCHITECTURE.md` lists
`SetProject` among the transitions and in the execution state diagram
(W, E0 and E1 back to themselves). It also says that a root chat's
project can change after creation, and that a moved chat's next turn
starts with a memory index update from the old project to the new one.

## Testing

- `make pre-commit` passes. `go test` passes for
`coderd/x/chatd/chatstate`, `coderd/database/dbauthz`
(`TestMethodTestSuite`) and the chatd memory tests.
- `TestSetFamilyProjectLocksRootFirst` fails with 55P03 when a subagent
row is locked first. `TestSetFamilyProjectKeepsTheCurrentProject` covers
the same-project move during a run. The transition matrix test gains
`SetProject` cases for W, E0 and E1.
- `TestMovedChatUsesNewProject` runs two real turns around a move: the
second turn uses the instructions and memories of the new project, and
`OutOfProject` fails without the memory index removal.

> Xum opened this PR on behalf of @ibetitsmike.
2026-10-19 12:38:17 +00:00
Garrett Delfosse 6ebefb8325 feat: add chat project sharing API (#30439)
API for chat project sharing (experiment `chat-projects`). Builds on
#30479 (merged).

Owners share a project with users, groups, or the whole organization
(the Everyone group, ID = organization ID). Chats in a shared project
stay private.

| Role    | Project and instructions | Memories   | ACL  | Delete |
|---------|--------------------------|------------|------|--------|
| `use`   | read                     | read/write | read | no     |
| `admin` | read/write               | read/write | edit | no     |

Only the owner, and administrators whose role grants delete, can delete
a project.

- **API:** `GET`/`PATCH .../chats/projects/{project}/acl`, same shapes
as `/chats/{chat}/acl`, plus SDK methods. Project responses include the
caller's `permissions` (`update`, `delete`, `share`); the UI hides edit
and delete accordingly.
- **Listing:** `GET /chats/projects` returns owned and shared projects.
Removes `GetChatProjectsByOwnerID`.
- **Chat access:** creating a chat in a project, and loading its memory
and instructions on each turn, requires the chat owner to own the
project or have an ACL grant (`chatd.ChatProjectUsableBy`). Role grants
don't count. If the grant is removed, existing chats lose memory and
instructions on their next turn.
- **Validation:** users and groups must be in the project's
organization; keys must be canonical UUIDs; callers can't change their
own role; the owner can't be added. Migration `000624` drops a user's
grants when they leave the organization, so rejoining doesn't restore
access.
- **Instructions:** readers who can't edit get 403 instead of 404, and
the UI hides editing.
- **Agent prompt:** memories are visible to and editable by sharees, so
the agent is told not to save secrets or personal details.

Known gap: project deletion still detaches chats, and a sharee's open
sidebar keeps showing them until reload. #30683 replaces this.

Not included: share notifications, `/acl/available`, sharing UI.

<details>
<summary>Decision log</summary>

- `use` grants memory read/write over HTTP, because agents in a sharee's
chats can already write memories. Making the agent read-only for sharees
was the alternative (@f0ssel chose sharees changing memories through
use).
- An administrator with `share` can grant themselves access and then use
the project. Accepted to keep access control simple (@f0ssel).
</details>

> Generated by Coder Agents.
2026-10-19 11:32:26 +00:00
Jaayden Halko 452654d07f test(coderd/database): use subtest context in TestProjectSkillQueries (#31011) 2026-10-19 19:05:27 +08:00
Jaayden Halko e289c6fe01 feat: add project skill API routes (#30827)
Adds experimental routes for managing a chat project's skills under
`/api/experimental/organizations/{organization}/chats/projects/{project}/skills`:
list, create, get, update (PATCH) and delete by skill name.

The handlers reuse the shared skill create and update path from the
personal and organization skill APIs. Reading needs `chat_project:read`
on the project and writing needs `chat_project:update`; callers without
access get 404, matching the project instructions routes. Duplicate
names and the 100-skill-per-project cap return 409. Writes are audited
as `chat_project_skill` with the skill name as the target. The SDK gains
the matching `ExperimentalClient` methods.

Stack (merge top to bottom):
- #30826
- #30827 👈
- #30838
- #30840
- #30854
- #30855
- #30856
- #30861
- #30880
- #30881
- #30882
- #30884
2026-10-18 13:16:23 +07:00
Jaayden Halko a379c849da feat: add project skill queries and authorization (#30826)
Adds the storage and authorization layer for chat project skills. The
`skills` table already allows a `project_id` owner (migration 000621);
this adds the queries for it and authorizes them through the parent chat
project, the same way project instructions work.

- New queries in `skills.sql`: insert, get by name, list metadata,
update (COALESCE on description, content and enabled), delete, and
`GetProjectSkillByIDForUpdate` for the shared update lock.
- dbauthz checks `chat_project:read` on the project for reads and
`chat_project:update` for writes. `authorizeChatProjectInstructions`
became `authorizeChatProjectConfig`, shared by instructions and skills.
`Skill.RBACObject()` still denies project rows so nothing reaches them
without going through the project.
- Audit: new `chat_project_skill` resource type (enum-only migration
000624), `AuditableChatProjectSkill`, and the usual
diff/request/table/FriendlyString/AuditFilter wiring. The API that emits
these entries comes in the next PR.

Stack (merge top to bottom):
- #30826 👈
- #30827
- #30838
- #30840
- #30854
- #30855
- #30856
- #30861
- #30880
- #30881
- #30882
- #30884
2026-10-18 12:58:39 +07:00
Bobby Ho 764e67bcff feat: expose skipped and canceled script outcomes in coderd and the Build Timeline (#30875)
feat: expose skipped and canceled script outcomes

Add `skipped` and `canceled` statuses across the Agent API, database,
SDK, and Build Timeline for startup scripts.

- Store and return both outcomes, preserving zero-duration timings,
  and reject unknown statuses. Rollbacks rewrite the new outcomes to
  `exit_failure` with exit code 255.
- Add typed legends, outcome-specific tooltips, and keyboard-accessible
  bars. Fix timeline crashes for `timed_out` and `pipes_left_open`.
- Bump the Agent API to v2.15 while agents continue using v2.14.
  Add API, migration, and Storybook coverage.

First of two PRs for PLAT-550. Agents do not emit the new statuses yet;
a follow-up after #30765 enables reporting. Shutdown scripts remain
outside the timeline, and agent health and badge rules are unchanged.

Refs: https://linear.app/codercom/issue/PLAT-550/phase-3-expose-skipped-and-canceled-script-outcomes
2026-10-16 15:55:28 +00:00
Asher 899d25a5b1 feat: add byte tracking to ssh and pty connections (#30389) 2026-10-16 00:21:40 -08:00
George K 6d34557d74 feat: add script ordering adoption telemetry (#30860)
Count `coder_script_order` declarations and literal `rule` blocks in
coderd from the existing START plan in
`CompletedJob.TemplateImport.Plan`,
including child modules, without multiplying counts by resource or
module instances. Persist these counts per template version on
successful import after START plan graph validation, without changing
the provisioner API.

Report a deployment-wide `script_order_summary` with the number of
templates using script ordering and total declaration and rule
counts. Include only active, successfully imported versions of
non-deleted templates so repeated imports do not inflate
adoption. Elect one reporting replica per snapshot period and report
explicit zeros when no templates use ordering.

The summary contains only aggregate counts, without template
identifiers or rule contents. Dynamic rule blocks are not counted
because Terraform omits them from configuration JSON.

Configuration-decoding failures are logged and skip updating the
counts without independently failing the import. Script-order summary
collection failures are logged and omit that summary while allowing
the rest of the snapshot to proceed.

Refs: https://linear.app/codercom/issue/PLAT-551
2026-10-15 15:42:33 -07:00
Steven Masley 03cba25a63 feat: list a workspace build's secrets in coder secret list (#30388)
`coder secret list` now shows the current workspace build's secrets next to your user secrets, and which user secrets they replace. Part of [RELAY-7](https://linear.app/codercom/issue/RELAY-7/agent-relay-deliver-provider-credentials-via-coder-secrets-instead-of).

**Stack** (design decisions in #30248)
1. #30248 Database schema, queries, and RBAC
2. #30249 Encryption at rest
3. #30250 Set secrets on workspace builds
4. #30251 Deliver secrets to the agent
5. #30252 Docs
6. **#30388 List a build's secrets (this PR)**
7. #30459 Audit a build's secrets

- New `GET /workspacebuilds/{workspacebuild}/secrets` lists a build's secrets (metadata only, never values) and which user secret each one replaces.
- Inside a workspace, `coder secret list` includes the current build's secrets. Outside a workspace, use `--workspace` or `--build-id`.
- The agent now exposes `CODER_WORKSPACE_BUILD_ID`, so the CLI knows which build it is running in.
- The agent manifest and the new endpoint use one shared resolver, so they always agree on which secret wins.
- The user secret API is unchanged.

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

- The manifest carries `workspace_build_id`; the agent sets it as `CODER_WORKSPACE_BUILD_ID`.
- `coderd/workspacesecrets.Resolve(user, build, policy)` decides which secret is delivered on each env var and file. The manifest no longer relies on entry order.
- `codersdk.WorkspaceSecret` embeds `UserSecret` and adds `workspace_build_id`, `ephemeral`, `cleared_at`, `env_replaces`, and `file_replaces`. Secrets whose values a later build cleared are still listed, with `cleared_at` set.
- CLI columns: `source` and `replaced by` are shown by default only when a build is included; `ephemeral` is opt-in. The CLI makes two requests: user secrets, then build secrets.
- Agents that predate the build ID fall back to `CODER_WORKSPACE_ID` and its latest build.

</details>

<details>
<summary>Decision log</summary>

(existing decision log, unchanged)

</details>

---

*Created by Coder Agents on behalf of @Emyrk.*
2026-10-15 17:40:00 -05:00
Steven Masley f25b6294b6 feat: set workspace secrets on workspace builds (#30250)
Lets clients set secrets when creating a workspace or a build. Secrets are saved with the build and are never sent to the provisioner. Part of [RELAY-7](https://linear.app/codercom/issue/RELAY-7/agent-relay-deliver-provider-credentials-via-coder-secrets-instead-of).

**Stack** (design decisions in #30248)
1. #30248 Database schema, queries, and RBAC
2. #30249 Encryption at rest
3. **#30250 Set secrets on workspace builds (this PR)**
4. #30251 Deliver secrets to the agent
5. #30252 Docs
6. #30388 List a build's secrets
7. #30459 Audit a build's secrets

- `secrets` is a new field on the create workspace and create build requests.
- Each secret is delivered as an env var, a file, or both. `ephemeral` secrets last for one build only.
- Secrets carry forward to every later build until a request replaces or removes them.
- A secret set on another secret's env var or file path replaces that secret entirely.
- A null `value` removes a secret, which needs `update` permission on the owner's workspace secrets.
- Deleting a workspace clears its secret values.

Secrets are stored but not delivered to the agent until #30251.

<details>
<summary>Example request</summary>

```json
POST /api/v2/workspaces/{workspace}/builds
{
  "transition": "start",
  "secrets": [
    { "name": "openai-key", "value": "sk-...", "env_name": "OPENAI_API_KEY" },
    { "name": "kubeconfig", "value": "apiVersion: v1 ...", "file_path": "~/.kube/config" },
    { "name": "session-jwt", "value": "eyJ...", "env_name": "SESSION_JWT", "ephemeral": true },
    { "name": "old-token", "value": null }
  ]
}
```

</details>

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

- Validation reuses the user secret field rules and the file path delivery policy. It rejects:
  - duplicate names, env names, or file paths within one request
  - a removal that also sets a target or `ephemeral`
- In the build transaction, the builder:
  1. copies the previous build's live, non-ephemeral secrets that the request does not name or take over
  2. inserts the requested secrets
  3. "Earlier builds' rows are cleared when the new build's job completes
- Copying and clearing run as `AsWorkspaceSecretManager`, so autostart and other system-initiated builds work without secret permissions. Requested secrets are authorized as the build's actor.
- Each row records `source` (`request` or `carry_forward`).
- Per-build limit violations return a 400.
- A completed delete in `provisionerdserver`, or an orphan delete in the builder, clears the workspace's values and keeps the rows.

</details>

---

*Created by Coder Agents on behalf of @Emyrk.*
2026-10-15 14:20:44 -05:00
Steven Masley 7aef110ca2 feat: add workspace secrets table, queries, and RBAC (#30248)
Adds the database table, queries, and permissions for workspace secrets: secrets set on a workspace build and delivered to the agent, never to Terraform. 

Each secret belongs to one workspace build. A new build copies the previous build's secrets, then the old copies lose their values but stay as history.
2026-10-15 13:32:40 -05:00
Garrett Delfosse 2b9fad8bba feat: add chat project sharing database layer (#30479)
Database and authorization layer for chat project sharing (experiment
`chat-projects`), part 1 of 2. #30439 adds the API on top of it. Nothing
here changes behavior: projects start with empty ACLs, no endpoint
writes them yet, and deleting a project still detaches its chats.
Project deletion is in #30683.

- **Schema:** migration `000621` adds `user_acl` and `group_acl` jsonb
columns to `chat_projects`, typed as `ChatACL` like chats, with GIN
indexes. It also adds the `chat_project:share` API key scope.
- **Queries:**
- `GetChatProjectsOwnedOrSharedWithUserID`: owned projects plus projects
with an ACL entry for the user directly, through a group, or through the
Everyone group (whose ID is the organization ID). Entries count only for
members of the project's organization. Keys match regardless of actions,
so dbauthz filters the rows.
- `IsChatProjectAccessibleByUserID`: whether a user owns a project or
holds a `read` or `*` grant on it, for code running under a system
subject. With sharing disabled, dbauthz answers owner-only.
- `GetChatProjectByIDForUpdate`: lets the ACL update read, modify, and
write in one transaction.
  - `UpdateChatProjectACLByID`.
- **RBAC:**
- New `share` action on `chat_project`. dbauthz requires it to update
the ACL.
- `ChatProject.RBACObject` attaches the project ACL. Memory objects get
a derived ACL: project `read` grants memory read, and `read` with
`update` grants memory create and delete. Wildcard grants count. Read
sharees change memories through their chats' agents, which #30439 gates.
Project instructions authorize through the same object, so a `read`
grant reads them and an `update` grant edits them.
- `--disable-chat-sharing` also covers projects: ACLs are ignored, the
ACL update is rejected, and roles deny `chat_project:share`. The help
text and chat sharing docs say so.
- **Audit:** `user_acl` and `group_acl` changes are tracked.
- **Tests:**
- Querier tests run the access queries both raw and through dbauthz
(owner, direct user, group, Everyone, wildcard grants, a member of
another org, grants without read).
- RBAC object and memory ACL tests, with an authorizer check, in both
sharing modes.
- dbauthz cases including the disabled guards, and the
`chat_project:share` negation in the roles test.

The generated TypeScript changes only add the new scope value.

> Generated by Coder Agents.
2026-10-14 10:42:00 -04:00
Ehab Younes f573ec3533 feat: report agent connection methods and app identities (#30467)
Agents now report how a connection was made and which app made it, so
the connection log records the method and the client app separately.

- Bump the Agent API to v2.13. `Connection` gains `connection_method`
  and `app_name`, and `type` is deprecated.
- SSH and reconnecting PTY handlers set the method themselves, so a
  client-supplied app name can't change it.
- SSH sessions report the app from `CODER_SSH_SESSION_TYPE`. Sessions
  without one count as SSH but report no app.
- Disconnect reports repeat the connect report's method, app, and
  client session ID.
- Coderd logs the reported method and normalized app name. For agents
  older than v2.13, it maps `type` as before.
- Coderd rejects unknown methods instead of logging them as SSH.

Closes CLIENT-1037
2026-10-14 13:54:17 +03:00
Bobby Ho 0ea3f16dd3 fix(coderd/database/migrations): renumber MCP OAuth2 resource migration to 000622 (#30722)
Main is red: #30625 and #30651 both added a migration numbered 000618,
and the
loader refuses to start with a duplicate number. See the failing run on
fa7cd62c8b.

- Rename 000618_mcp_server_oauth2_resource.{up,down}.sql to 000622, the
next
  free number after 000621_skills.
- No schema change; dump.sql regenerates identically.
2026-10-13 23:08:37 +00:00
Michael Suchacz fa7cd62c8b feat: send RFC 8707 resource parameter on MCP server OAuth2 flows (#30651)
Coder's MCP OAuth2 client never sent the RFC 8707 `resource` parameter,
which the MCP authorization spec requires on authorization and token
requests. Authorization servers that enforce it cannot be connected:
with Dagster+, discovery and dynamic client registration succeed, then
the authorize request fails with "MCP OAuth requests must include
resource".

MCP server configs get an `oauth2_resource` field (migration 000618,
admin-only in API responses, audited). Coder sends it on the authorize
redirect, the authorization-code exchange, and token refresh. Refresh
now goes through `clientcredentials` with `grant_type=refresh_token`,
because `oauth2.Config` cannot add parameters to refresh requests; it
keeps the same auth-style detection and error types, and adds no
dependency.

Auto-discovery fills the field with the most specific Protected Resource
Metadata `resource` that covers the server URL, stored as advertised. As
in the MCP SDKs, a resource covers URLs on its origin at or under its
path, so `/mcp` covers `/mcp/` but `/mcp/` does not cover `/mcp`. URLs
compare in RFC 3986 normal form (default ports, percent-encodings, dot
segments, and an empty path as `/`), ports ignore leading zeros, IPv6
literals compare by address, and `%2F` stays inside its segment. Unlike
the SDKs, a resource with a query, even an empty one, only covers a
server URL with the same query. When none covers it, Coder uses the
server URL with the scheme and host lowercased and user information and
fragment removed, or sends no resource when that is not a valid resource
indicator. An explicit value in the create request wins.

Manual configurations send no `resource` unless an admin sets one,
because some providers reject it (Microsoft Entra ID v2 fails with
AADSTS901002). Values must be absolute RFC 3986 URIs without a fragment,
user information, or an IPv6 zone identifier.

The resource binds grants like the other OAuth settings: changing it
deletes stored user tokens, and a callback whose config resource changed
after the user started connecting returns 409 before the code is
redeemed. The connect step records the resource in a callback-scoped
cookie for that check.

Discovery only runs when a configuration is created, so existing
auto-discovered configurations, including the Dagster+ one that prompted
this, keep an empty resource until an admin sets the new "Resource URL"
field or re-creates the server.

![MCP server edit form showing the Resource URL field filled by
auto-discovery](https://github.com/user-attachments/assets/319bc8a0-f6e1-4b7d-8878-8e341c94e590)

## Testing

- Go: discovery (most specific match, path and trailing-slash
boundaries, query matching including empty queries, equivalent ports,
IPv6 literals, escapes, and dot segments including escaped ones, the
normalized fallback and an invalid one, explicit override), validation
(URI characters, brackets, IP literals, user information, trimming),
grant deletion on change, clearing on an auth type change, redaction for
members, refresh with and without a resource, the 409 when the resource
changes mid-flow (the authorization server never receives the code), and
a connect and callback flow against an authorization server that rejects
requests without the resource. Each new test was checked to fail when
the behavior it covers is broken, except two malformed IP literal hosts
that Go's URL parser already rejects.
- Vitest: the form round-trips the stored value, and clearing the field
sends `""` on update.
- Remote UAT through Coder Agents
([chat](https://dogfood.cdr.dev/agents/58695d5b-2112-400b-88f7-ab64c678f756))
on a test deployment with loopback MCP and authorization server mocks:
- Round 1 on 2b875068e0: literal verdict FAIL. When the resource
changed between connect and callback, the callback redeemed the code for
the new resource and stored a grant. The 409 binding fixes it.
- Round 2 on 4b54ffb5f9: PASS. The mid-flow change returned 409 with no
token request in every variant (strict and lenient servers, set,
cleared, abandoned flows, parallel configs); auto-discovery, connect,
refresh, validation, and the docs matched. Minor observations: the
binding cookie is an unkeyed digest, so a member who edits their own
HttpOnly cookie can redeem a code with the current resource, which a
fresh connect also gives them; other settings changed mid-flow still
reach the token endpoint, as on main; the 409 popup shows raw JSON; the
UI does not warn that a resource change disconnects users; validation
toasts omit the server's detail.
- Round 3 on the final head 2d6abfac49: PASS. The head moved during the
round, so it ran in four parts, the last on 2d6abfac49. That part
covered the matching rules (queries including empty ones, default and
zero-padded ports, `%2F`, escaped dot segments, trailing slashes, IPv6
literals), the invalid fallback (stored empty, no `resource` sent),
every validation value through the API and the UI, and the discovery,
connect, refresh, and mid-flow 409 flows. The equivalent-encoding cases
last ran live on 5c23fe00dca; unit tests cover them on the final head.
Observation: the path-aware metadata URL is built from the decoded path,
so `/mcp%2Ftenant` fetches metadata for `/mcp/tenant`, as on main.
- `frontend-review` and the eight-role `coder-frontend-review` ran on
the `site/` diff; every role passes after the label, help text, row
layout, API doc wording, and form test fixes. Declined: linking field
descriptions with `aria-describedby`, because the form's shared `Field`
primitive renders every description without one, so it belongs in a
change to all fields.

## Review record

| Review | Commit | Findings | Disposition |
|---|---|---|---|
| [Codex round
1](https://github.com/coder/coder/pull/30651#pullrequestreview-5492085359)
| 4b54ffb5f9 | P2: a resource with another query was treated as
covering the server URL | Reproduced with a failing test, fixed in
bf0879560c, resolved |
| [Codex round
2](https://github.com/coder/coder/pull/30651#pullrequestreview-5492305363)
| bf0879560c | P2: explicit default ports; P2: `%2F` treated as a
segment boundary; P2: `url.Parse` accepts spaces and invalid escapes |
Each reproduced with a failing test, fixed in 16a1d783fa, resolved |
| [Codex round
3](https://github.com/coder/coder/pull/30651#pullrequestreview-5492677517)
| 16a1d783fa | P1: docs wrapped mid-sentence (DOC1); P2: equivalent
percent-encodings compared unequal; P2: brackets accepted outside an IP
literal host | Fixed in 8d617f0528 (the P2s reproduced with failing
tests; the same pass also covers the query, dot segment, and user
information cases), resolved |
| [Codex round
4](https://github.com/coder/coder/pull/30651#pullrequestreview-5493033885)
| 8d617f0528 | P2: the server URL fallback was not validated; P2:
escaped dot segments were compared before decoding | Each reproduced
with a failing test, fixed in 5c23fe00dc, resolved |
| [Codex round
5](https://github.com/coder/coder/pull/30651#pullrequestreview-5493342889)
| 5c23fe00dc | P2: an empty query (`?`) was treated as no query |
Reproduced with a failing test, fixed in 72c019ff5e, resolved |
| [Codex round
6](https://github.com/coder/coder/pull/30651#pullrequestreview-5493524846)
| 72c019ff5e | P2: `/mcp/` covered `/mcp`; P2: equivalent IPv6 literals
compared unequal; P2: malformed IP literals accepted | The first two
reproduced with failing tests. Go 1.26.5 `url.Parse` already rejects the
hosts named in the third, but it accepted an IPv6 zone identifier. All
fixed in ef2df2c4e7, resolved |
| [Codex round
7](https://github.com/coder/coder/pull/30651#pullrequestreview-5493668131)
| ef2df2c4e7 | P2: leading zeros in a port made equal origins differ |
Reproduced with a failing test, fixed in 2d6abfac49, resolved |
| [Codex round
8](https://github.com/coder/coder/pull/30651#issuecomment-6128743972) |
2d6abfac49 | None: "Didn't find any major issues" | Clean verdict on
the current head |

> Xum opened this PR on behalf of @ibetitsmike.
2026-10-13 23:03:08 +02:00