You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
feat: Settings shows the attendant routine's setup and prompt - #124
Settings → Call an agent gets collapsible setup steps and the routine prompt (with a Copy button). The prompt lives in the repo, filled with this server's URLs, a test keeps it in step with the fired payload, and it ends the payload-less first turn without notifying.
…ompt to copy
The prompt ships with the server, filled with this server's URLs, and ends a
turn quietly when the payload has not arrived yet.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Bwp5QefjsjMFeq6CK6cT6w
The reason will be displayed to describe this comment to others. Learn more.
No issues found. Checked that the prompt placeholders resolve to the mounted /mcp and /live/await routes, that build.ts ships the prompt into dist, that the Settings CSP allows only same-origin scripts, and that the drift test compares the prompt's payload line with an actual fire.
The reason will be displayed to describe this comment to others. Learn more.
🟡 Changes recommended
Patch sessions are directed to fetch an unpushed tree from origin despite lacking access to fallback retrieval tools, leaving routine calls without a working retrieval path or a clear rejection.
Tour instructions omit required arguments and reference deleted files
packages/server/src/routine-prompt.md:59
The raw HTTP helper does not retrieve tool schemas, yet this instruction omits the required tour argument for tour_step and tour_done, and the nonempty body for tour_step. It also requests steps for deleted files, which tour_step rejects because it reads files at the session head (mcp.ts:518–555). Include complete call shapes, restrict steps to files present at the head, and describe deletions in the tour body.
The reason will be displayed to describe this comment to others. Learn more.
🔵 Needs a closer look
Human assessment is needed because an unpublished review finding raises a concern about whether the comment-tool example guides reviewers to anchor removed-code findings to the correct diff side.
The reason will be displayed to describe this comment to others. Learn more.
🔵 Needs a closer look
Human assessment is needed because commands may use the wrong checkout in multi-repository setups, causing fetch failures and leaving reviews or questions unhandled.
The reason will be displayed to describe this comment to others. Learn more.
🟢 Approval recommended
No active review issues are reported for the routine setup guidance and patch-session safeguards.
No open issues 🎉
Review effort: Balanced
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Settings → Call an agent gets collapsible setup steps and the routine prompt (with a Copy button). The prompt lives in the repo, filled with this server's URLs, a test keeps it in step with the fired payload, and it ends the payload-less first turn without notifying.
🤖 Generated with Claude Code
https://claude.ai/code/session_01Bwp5QefjsjMFeq6CK6cT6w