.agents/skills/rstack-eco-ci-debug/references/automation-daily-triage.md
Use this reference for recurring/daily Rspack eco-ci automation runs and latest-status monitoring tasks.
Read this file before inspecting runs when the request:
/goal.For a single job, PR, suite, or local reproduction request, use the local quick path in SKILL.md instead.
rspack-ecosystem-ci-from-commit run and the previous completed run. Do not use in-progress runs. Prefer scripts/rspack-status.sh --repo <ecosystem-ci-path> from the skill directory when a local rstack-ecosystem-ci checkout is available, or pass a local rspack.json snapshot path directly.reused: same suite and same failure signature as memory/current sampled log.new-suite: suite is failing now but was not failing in the previous completed run.new-signature: suite was already failing, but the current failure signature changed.recovered: suite was failing previously but is no longer current; mention only in the summary.new-suite or new-signature, run Phase 1. Run Phase 2 only when Phase 1 finds a non-flaky candidate Rspack PR or version window with enough evidence.pr-report-comment.md and comment only when its guardrails are satisfied. Otherwise include no PR comment: <reason> in the report.Use the triage report contract from SKILL.md, with these daily-specific fields:
Scope: latest completed status.
Current run/job: <latest completed run-url>, testing <sha> — "<commit msg>".
Comparison baseline: previous completed run <run-url>, testing <sha>.
Failing suites in scope: <suite list>.
Delta: <new-suite/new-signature/recovered/unchanged summary>.
PR comments: <posted links or "none: <reason>">.
Every currently failing suite still gets its own short section. Do not include recovered suites as failing-suite sections.