fix(ci): render llama canary agent output as text - #1867
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Advanced Run ID: 📒 Files selected for processing (3)
Included review availability: Your plan provides up to 8 included reviews per hour; 7 remain after this review. 📝 WalkthroughWalkthroughChangesCanary output contract
Priority: ⬇️ Low Estimated code review effort: 1 (Trivial) | ~5 minutes Change: Bug fix Merge Risk: ⚪ Minimal · up to The wrapper now produces readable live and retained logs without changing repair status handling or verification gates, so the change is mergeable. 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Full details: Docstring CoverageExplanation Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 2 functions across 2 files. (1 skipped: 1 unsupported.)
✨ Finishing Touches 💡 1📝 Generate docstrings 💡
🧪 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 |
Original problem
The changed-pin llama canary asked Goose for
stream-jsonand copied the event stream directly into the live Actions log. Since Goose emits assistant thinking and text as small streaming messages, a short sentence appeared as many full JSON objects and was difficult to follow while the canary ran.Diagnostics
The harness had no structured-output consumer:
scripts/llama-canary-agent-repair.shsent--output-format stream-jsonto Goose and piped that stdout directly totee. Goose already provides a nativetextrenderer for assistant text, thinking, tool calls, tool results, and errors.Fix
Use Goose's native
textoutput mode for the canary agent session. The existing live output and retainedagent.lognow receive Goose-formatted human-readable output, while session naming, resume behavior, provider/model selection, credentials, timeouts, verification gates, and publication controls remain unchanged.The harness contract test now fixes the readable output mode, and the CI inventory documents the behavior.
Validation
just ci-validate— 1,461 tests passed, 9 skipped; Actionlint and all repository consistency gates passedpython3 -m unittest scripts.tests.test_llama_canary_agent_repair_contract— 20 passedjust ci-shellcheck scripts/llama-canary-agent-repair.shSummary by CodeRabbit
Documentation
Bug Fixes
Tests