Repository navigation
Make main green for 0.1.122 (2): show_html's dashboard sentence, the boxed Chrome and the team browser stub off Windows - #3304
Conversation
…xed Chrome and the team browser stub off Windows - server/drivers/agents-call-cards.test.ts: show_html's description now ends by pointing a dashboard at the bundled data-dashboards skill (#3279); the test's end-anchored match includes that sentence. - server/bot-box/box.test.ts: "the box a bot's Chrome runs in" starts the box, whose gates are unix sockets Windows refuses at a file path (EACCES); that test skips on win32. Its "no box off a Cloud home" check, what Windows gets, is its own test and runs everywhere. - server/cloud-team-browser.e2e.test.ts: the agent-browser stand-in is a sh script, which Windows cannot run as a program, so no command reached its log; that one test skips on win32 (a Team Cloud is a Linux machine). The file's other checks still run there. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
There was a problem hiding this comment.
If your organization's extra usage balance is empty, an organization admin can add extra usage credits at claude.ai/admin-settings/usage. If its monthly spend limit was reached, an admin can raise it on the same page. If neither applies, contact Anthropic support.
Once extra usage is available, someone with write access to this repository can comment @claude review on this pull request to trigger a review.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info
📝 Walkthrough
Merge Risk: ⚪ Minimal · up to
|
| Check name | Status | Explanation |
|---|---|---|
| Title check |  |
The title clearly identifies the main changes: the show_html dashboard sentence and Windows-specific test handling. It is longer than ideal but remains specific and related to the changeset. |
| Description check | |
The description explains what changed, why the changes were needed, how the tests were verified, and which unrelated failures were treated as flakes. It does not use the template headings or checklist… |
| Docstring Coverage | |
No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 3… |
| Linked Issues check | |
Check skipped because no linked issues were found for this pull request. |
| Out of Scope Changes check | |
Check skipped because no linked issues were found for this pull request. |
Full details: Description check
-
Explanation
The description explains what changed, why the changes were needed, how the tests were verified, and which unrelated failures were treated as flakes. It does not use the template headings or checklist, but it is sufficiently complete and relevant.
Full details: Docstring Coverage
-
Explanation
No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 3 files.
✨ Finishing Touches
-
📝 Generate docstrings
-
- Commit to this branch
- Create a new PR
🧪 Generate unit tests (beta)
-
- Commit to this branch
- Create a new PR
-
- Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.
Comment @coderabbitai help to get the list of available commands.
Main went red again at
dbb73a73a, after #3273, #3275, #3276, #3277, #3278, #3280, #3284, #3285 and #3286 landed on top of #3281. The re-pinned 0.1.122 release then failed its gate on CI run 39716552977. This PR changes no product code.server/drivers/agents-call-cards.test.ts, "the chat-app tools are in every catalog…" (all OSes)show_html's description now ends "For a data dashboard, follow the data-dashboards skill." (#3279). The test's end-anchored regex now includes that sentence. The tool order was not the problem.server/bot-box/box.test.ts, "the box a bot's Chrome runs in" (Windows)it.skipIf(win32). Its "no box off a Cloud home" check, which is what Windows gets, is now a test of its own and runs everywhere.server/cloud-team-browser.e2e.test.ts, "drives each member's private chat…" (Windows)#!/bin/shscript. Windows can't run it as a program, so no command ever reached its log (expected [] to include 'bot-…'). A Team Cloud is a Linux machine. Only this test skips on win32; the file's other checks (live-view audience, phone pairing) still run there.server/team-move.e2e.test.ts(Ubuntu shard 3)dbb73a73a. One occurrence; it's rerun once below if it recurs.server/index.test.ts, "parks the turn at the wait ceiling…" and "lists sanitized cloud computers…" (Windows shard 4)index.test.tstest fails each time on a loaded Windows or Linux shard. Neither file changed in a way that touches these paths.I refreshed the shard weights from the three latest Windows runs and then reverted. It would even out Windows (max 2019 → 1866 s) but push Linux's slowest shard from 1427 to 1499 s, and Linux is the shard nearest its cap.
Heads-up: in the gate run, Linux shards took 24.6–27.9 min against a 30-minute cap, and Windows shard 1 took 41.4 min against 45.
scripts/ci-workflow.test.tspins those caps, so I left them alone.Local (macOS):
agents-call-cards,box,cloud-team-browser.e2eandteam-move.e2epass. Lint and the server typecheck pass too.🤖 Generated with Claude Code
Summary by CodeRabbit