fix(ci): drop stale --no-stdio flag from mcp publish startup#75
fix(ci): drop stale --no-stdio flag from mcp publish startup#75sergiofilhowz merged 1 commit intomainfrom
Conversation
The mcp/v0.5.1 publish failed because `_publish-registry.yml` started the worker with `cargo run --bin mcp -- --no-stdio`, but mcp's CLI no longer accepts that flag (mcp/src/main.rs only declares `--config`, `--url`, `--manifest`). cargo exited with `error: unexpected argument '--no-stdio' found` before the worker could connect, the 2s `kill -0` post-spawn check passed because cargo itself was still draining, and `collect_worker_interface` then waited 120s for a worker that never registered before raising `ValueError: could not match worker 'mcp'`. The flag was added in #67 specifically to keep mcp from binding stdio in the publish job; that concern is moot now (default mcp run is fine for interface collection). Drop the special case so mcp goes through the same `cargo run --bin mcp --` path as every other rust worker.
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
💤 Files with no reviewable changes (1)
📝 WalkthroughWalkthroughThis PR removes the special-case conditional that appended ChangesWorkflow Simplification
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~2 minutes Possibly related PRs
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
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 |
Summary
mcp/v0.5.1publish failed (job) with the worker dying immediately on startup:_publish-registry.yml:141-143had a worker-specific special case from #67:mcp's CLI no longer accepts
--no-stdio—mcp/src/main.rs:25-34only declares--config,--url,--manifest. cargo exited before the worker could connect, the 2skill -0check after spawn still passed (cargo itself was technically still alive), andcollect_worker_interfacethen waited the full 120s before raisingValueError: could not match worker 'mcp'.Fix
Drop the special case. mcp's default startup is fine for interface collection — the original concern that motivated
--no-stdiodoesn't apply in CI.Side benefit: removes the worker-name-in-shared-CI anti-pattern flagged in the original review of #67. If a worker ever needs CI-specific startup args again, declare them in
iii.worker.yamlinstead.Test plan
create-tag.ymlwithworker=mcp, bump=patch, tag=next.Start local worker for interface collectionstep showscargo run --bin mcp --(no--no-stdioappended).collect_worker_interfacefinds it, publish completes with HTTP 200.Summary by CodeRabbit