fixes https://github.com/coder/internal/issues/1203
The matcher I wrote for TestAgentConnectionMonitor tested that `last_disconnected_at` was strictly _after_ the start of the test to ensure it was updated.
This is too strict of a test because Windows in particular doesn't have high-resolution timers, so it's entirely possible to get the exact same timestamp from subsequent calls to `time.Now()`. This PR switches the test to _not before_ to cover this case. The results are just as valid because we always initialize the `last_disconnected_at` to something well before the test starts.
A simple typo fix to the help text
stidin > stdin
```
➜ coder git:(org_role_fix) ✗ coder organizations roles create -h
coder v2.29.1+59cdd7e
USAGE:
coder organizations roles create [flags] <role_name>
Create a new organization custom role
- Run with an input.json file:
$ coder organization -O <organization_name> roles create --stidin < role.json
```
Add the AgentAPI changes to support the feature that transmits boundary
logs from workspaces to coderd via the agent API for eventual re-emission to
stderr. The API handlers are stubs for now because I'm trying to land
this feature from multiple smaller PRs.
High level architecture:
- Boundary records resource access in batches and sends proto message to
agent
- Agent proxies messages to coderd **(captured by the API changes in
this PR)**
- coderd re-emits logs to stderr
RFC:
https://www.notion.so/coderhq/Agent-Boundary-Logs-2afd579be59280f29629fc9823ac41ba
Related to
[`internal#1139`](https://github.com/coder/internal/issues/1139)
Continuation of #21074
This implements some RBAC role specificity for `dbpurge`, ensuring that
we follow the least-privileged model for removing data from the
database. It is specified as following.
```go
Site: rbac.Permissions(map[string][]policy.Action{
// DeleteOldWorkspaceAgentLogs
// DeleteOldWorkspaceAgentStats
// DeleteOldProvisionerDaemons
// DeleteOldTelemetryLocks
// DeleteOldAuditLogConnectionEvents
// DeleteOldConnectionLogs
rbac.ResourceSystem.Type: {policy.ActionDelete},
// DeleteOldNotificationMessages
rbac.ResourceNotificationMessage.Type: {policy.ActionDelete},
// ExpirePrebuildsAPIKeys
// DeleteExpiredAPIKeys
rbac.ResourceApiKey.Type: {policy.ActionDelete},
// DeleteOldAIBridgeRecords
rbac.ResourceAibridgeInterception.Type: {policy.ActionDelete},
}),
```
| Position | Pull-request |
| -------- | ------------ |
| | [feat: add prometheus observability metrics for
`dbpurge`](https://github.com/coder/coder/pull/21074) |
| ✅ | [feat: add rbac specificity for
`dbpurge`](https://github.com/coder/coder/pull/21088) |
Related to
[`internal#1139`](https://github.com/coder/internal/issues/1139)
This implements some prometheus metrics for records being removed from
the database. Currently we're tracking the following fields being
removed from the DB by this. They're viewable in the
`/api/v2/debug/metrics` endpoint.
* `expired_api_keys`
* `aibridge_records`
* `connection_logs`
* `duration`
```
# HELP coderd_dbpurge_iteration_duration_seconds Duration of each dbpurge iteration in seconds.
# TYPE coderd_dbpurge_iteration_duration_seconds histogram
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="1"} 1
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="5"} 1
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="10"} 1
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="30"} 1
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="60"} 1
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="300"} 1
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="600"} 1
coderd_dbpurge_iteration_duration_seconds_bucket{success="true",le="+Inf"} 1
coderd_dbpurge_iteration_duration_seconds_sum{success="true"} 0.014787814
coderd_dbpurge_iteration_duration_seconds_count{success="true"} 1
# HELP coderd_dbpurge_records_purged_total Total number of records purged by type.
# TYPE coderd_dbpurge_records_purged_total counter
coderd_dbpurge_records_purged_total{record_type="aibridge_records"} 0
coderd_dbpurge_records_purged_total{record_type="audit_logs"} 0
coderd_dbpurge_records_purged_total{record_type="connection_logs"} 0
coderd_dbpurge_records_purged_total{record_type="expired_api_keys"} 0
coderd_dbpurge_records_purged_total{record_type="workspace_agent_logs"} 0
```
| Position | Pull-request |
| -------- | ------------ |
| ✅ | [feat: add prometheus observability metrics for
`dbpurge`](https://github.com/coder/coder/pull/21074) |
| | [feat: add rbac specificity for
`dbpurge`](https://github.com/coder/coder/pull/21088) |
Relates to #20925
This PR modifies TaskPage to update an outdated workspace instead of
starting it. Before, starting an outdated workspace where the template
required the active version would fail with the error "cannot use
non-active version: rbac: forbidden".
For the case of a dormant workspace, I deemed it safe enough to simply
unset dormancy on an attempted start (ref:
https://github.com/coder/coder/pull/21306). However, automatically
updating a workspace is a more risky option, so I instead elected to
give the user the option of updating their workspace using the existing
tooltip.
**Note:** I made a change to the `WorkspaceOutdatedTooltip` components
to allow it to have children so that the tooltip could trigger over a
wider element instead of just the info icon.
```
<🤖 AI Disclaimer>I got some help from Gemini 3 Flash in "Ask" mode.</🤖 AI Disclaimer>
```
While scale testing, I noticed that our load generators send basically
all requests to a single Coderd instance.
e.g.
This is because our scale test commands create all `Runner`s using the
same codersdk Client, which means they share an underlying HTTP client.
With HTTP/2 a single TCP session can multiplex many different HTTP
requests (including websockets). So, it creates a single TCP connection
to a single coderd, and then sends all the requests down the one TCP
connections.
This PR modifies the `exp scaletest` load generator commands to create
an independent HTTP client per `Runner`. This means that each runner
will create its own TCP connection. This should help spread the load and
make a more realistic test, because in a real deployment, scaled out
load will be coming over different TCP connections.
follows on from https://github.com/coder/registry/pull/616
As part of
https://registry.terraform.io/providers/hashicorp/kubernetes/latest/docs/guides/v3-upgrade-guide
various resources change from non versioned to versioned. This PR
changes the Coder authored templates to versioned resources.
- Updated kubernetes_persistent_volume_claim to
kubernetes_persistent_volume_claim_v1
- Updated kubernetes_deployment to kubernetes_deployment_v1
- Updated kubernetes_pod to kubernetes_pod_v1
- Updated kubernetes_secret to kubernetes_secret_v1
- Updated all resource references and dependencies
Closes#21052 and is related to #21218
This pull request implements a bunch of changes the team requested to
improve the overall user experience when navigating to the Request Logs
page. Since these are mostly UI changes, screenshots do a much better
job of showing what's different than me trying to explain everything in
words, so I've included them below.
> [!WARNING]
> More globally we've changed how we're rendering `<PaginationHeader
/>`, therefore we've renamed it to `<PaginationAmount />`. This will
effect multiple layouts however its a global improvement we're
intentionally looking to make.
### Preview
### Changes
1. Date formatting set to a standardised way including the month as a
`MMM` (i.e `Dec`)
2. Names and avatars are now rendered larger, furthermore they render
the `.name` field falling back on `.username`.
3. Font-size increased from `text-xs` to `text-sm` to increase overall
legibility globally.
4. Initiator is now rendered inside of a `<Badge />` and gives longer
names a chance to render.
5. Models are now rendered in a `<Badge />` alongside their icons.
6. Input/Output tokens are in a shared `<Badge />`-group alongside
truncation (when necessary).
7. Badge from `5` is rendered with the proper name being visible on
hover (for use if truncated).
8. Move globally the `Showing X of X` counts to the bottom of the tables
(effects other components outside of AI Bridge).
9. Icons are now included within the dropdown for the `Provider` filter.
10. Truncation of elements is now handled so larger prompts don't cause
the table to require further scrolling.
11. New merged Input/Output `<dt>` element with the badge
| Position | Pull-request |
| -------- | ------------ |
| ✅ | [fix: improve AI Bridge request logs
UI/UX](https://github.com/coder/coder/pull/21252) |
| | [feat: add AI Bridge request logs model
filter](https://github.com/coder/coder/pull/21259) |
| | [chore!: promote AIBridge from
`ExperimentalHandler`](https://github.com/coder/coder/pull/21278) |
| | [feat: implement request log collapsing prompt (`<RequestLogsPrompt
/>`)](https://github.com/coder/coder/pull/21313) |
---------
Co-authored-by: ケイラ <mckayla@hey.com>
Relates to #20925
This PR modifies the `postWorkspaceBuild` handler to automatically unset
dormancy on a workspace when a start transition is requested.
Previously, the client was responsible for unsetting the dormancy on the
workspace prior to posting a workspace build.
Adds `--max-failures` flag to `coder exp scaletest create-workspaces` so that we can tolerate a few failures without failing the command.
When running our scale test infra, we create Kubernetes Jobs to create the initial cluster workspaces, then we have load-generation jobs that depend on them. At high scale, it's kind of expected that some of the requests will fail: even with 99.9% success, you still expect one failure per 1000. It's useful to be able to carry on with the scale test anyway and proceed to traffic generation.
Relates to #20925
This PR expands the test coverage of `enterprise/coderd/TestWorkspaceBuild` to also exercise the `postWorkspaceBuilds` handler. Previously, it only exercised the `createWorkspace` handler.
I noticed while looking at scale test metrics that we don't always
report a useful path in the API request metrics.
There are a lot of requests with path `/*`. I chased this problem to the
workspace proxy, where we mount a the proxy router as a child of a
"root" router to support some high level endpoints like `latency-check`.
Because we query the path from the Chi route context in the prometheus
middleware _before_ the request is actually handled, we can have a
partially resolved pattern match only corresponding to the root router.
The fix is to always re-resolve the path, rather than accept a partially
resolved path.
This PR separate the data retrieval for workspace sharing ACL into a
custom hook and creates a separate form component. This is in
preparation for reusing the workspace sharing form from a new share
button on the workspace page.
fixes#21303
Update user last_seen_at when we mark them active on login. This prevents a narrow race where they can be re-marked dormant and fail to log in.
Closes#21217
This pull-request traverses the `token_usages.metadata[...]` fields and
merges them into a single consumable object so that we're able discern
information about the metadata within the token usages at a glance. Its
not the be-all end-all implementation of this feature but its a stepping
stone in order to render more useful data to the frontend.
These are currently mergable because they only contain `number` based
fields. When it encounters something within the object that can't be
merged (minus empty objects `{}`) it will simply return them as an
array.
### Preview
<img width="2682" height="1360" alt="CleanShot 2025-11-28 at 15 30
09@2x"
src="https://github.com/user-attachments/assets/e07e6515-4b8e-4169-841c-38fd83c434f9"
/>
### Logic breakdown
<img width="914" height="1016" alt="CleanShot 2025-11-28 at 15 11 13@2x"
src="https://github.com/user-attachments/assets/34b78fe1-3b58-4b78-a552-028ea5a88dc4"
/>
This removes the deprecated AITaskPromptParameterName constant and all
backward compatibility code that was added for v2.28.
- Remove AITaskPromptParameterName constant from codersdk/aitasks.go
- Remove backward compatibility code in coderd/aitasks.go that populated
the "AI Prompt" parameter for templates that defined it
- Remove the backward compatibility test (OK AIPromptBackCompat)
- Update dbfake to no longer set the AI Prompt parameter
- Remove AITaskPromptParameterName from frontend TypeScript types
- Remove preset prompt read-only feature from TaskPrompt component
- Update docs to reflect that pre-2.28 definition is no longer supported
Task prompts are now exclusively stored in the tasks.prompt database
column, as introduced in the migration that added the tasks table.
BREAKING CHANGE: SFTP/SCP now respects the agent's configured directory.
If your workspace agent has a custom `dir` configured in Terraform, SFTP
and SCP connections will now land there instead of `$HOME`. Previously,
only SSH and rsync respected this setting, which caused confusing behavior
where `scp file.txt coder:.` and `rsync file.txt coder:.` would put files
in different places. If you have scripts that relied on SFTP/SCP always
using `$HOME` regardless of agent configuration, you may need to use
explicit paths instead.
Add a fallback step that installs helm via apt from Buildkite's
repository if the primary `azure/setup-helm` action fails. This makes CI
resilient to `get.helm.sh` outages while preserving the caching benefits
of the action when it works.
## Changes
- Primary: `azure/setup-helm` with `continue-on-error: true`
- Fallback: apt install from Buildkite's helm-debian repository (only
runs if primary fails)
The fallback uses GPG-signed packages from the [official Helm apt
repository](https://helm.sh/docs/intro/install/#from-apt-debianubuntu)
hosted by Buildkite.
## Background
On Oct 29, 2025, `get.helm.sh` experienced an outage that broke CI. A
workaround was applied (#20552) but later reverted. This PR makes the
workaround automatic - it only kicks in when needed.
Fixes: https://github.com/coder/internal/issues/1109
Fix a typo in the MCP documentation where "seems" should be "sees":
> These inner loops are not relayed back to the client; all it **sees**
is the result of this loop.
Found while reading the docs.
**Breaking Change:** Existing oauth apps might now use PKCE. If an
unknown IdP type was being used, and it does not support PKCE, it will
break.
To fix, set the PKCE methods on the external auth to `none`
```
export CODER_EXTERNAL_AUTH_1_PKCE_METHODS=none
```
Provisioner steps broken into smaller granular actions.
Changes:
- `ExtractArchive` moved to `init` request (was in `configure`)
- Writing `tfstate` moved to `plan` (was in `configure`)
- Moved most plan/apply outputs to `GraphComplete`
closes: https://github.com/coder/internal/issues/858
Similar to https://github.com/coder/coder/pull/19375, this one uses
system permissions for fetching actual user and group data.
Modifies the `workspaces_expanded` view to fetch the required data; this way it's made available to all code paths that make use of it.
Also fixes a bug in a test helper function that can result in `null` being saved to the DB for `user_acl` or `group_acl` and break tests; a defensive check constraint that prevents this is worth a PR, e.g:
`ALTER TABLE workspaces
ADD CONSTRAINT group_acl_is_object CHECK (jsonb_typeof(group_acl) = 'object');`
Also adds missing `OwnerName` in `ConvertWorkspaceRows`.
This change ensures keyring tests that utilize the real OS keyring use
credentials that are isolated by process ID so that parallel test processes
do not access the same credentials.
https://github.com/coder/internal/issues/1192
closes#19974
This PR replaces all remaining MUI tooltips with our shadcn tooltip
component, and uncomments the linter restriction on importing the MUI
tooltip
I noticed we have a defer to make sure all log output is captured by the
async log handling routine.
```golang
defer func() {
_ = outWriter.Close()
_ = errWriter.Close()
<-doneOut
<-doneErr
}()
```
But `e.timings.aggregate()` compiles the **current** parsed logs. I was
getting some race conditions in testing if the logs did not completely
parse before `aggregate` was called.
So now `aggregate` happens outside the cmd exec function, to make sure
it is completed before logs are accumulated.
Bumps rust from `5218a2b` to `0d8bf26`.
[](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)
Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.
[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)
---
<details>
<summary>Dependabot commands and options</summary>
<br />
You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
</details>
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
