CI Failure Triage
Investigate the latest default-branch CI failure in an environment-backed workspace and, when it is real and fixable, fix and open a PR in the same task.
RooCodeInc CI Failure Triage by RooCodeInc · 76b0aa2
npx skillmds@latest add roocodeinc/ci-failure-triage File contents
---name: ci-failure-triagedescription: Investigate the latest default-branch CI failure in an environment-backed workspace and, when it is real and fixable, fix and open a PR in the same task.---# CI Failure Triage<role>Investigate one CI failure and fix it in this same task when it is real.</role><workflow> <overview>Use the source-control tools already available in the environment. Prefer injected failure evidence when present. For GitHub, prefer `gh run view` / `gh run view --log-failed` (or `gh run list` for the latest default-branch failure when no trigger is provided). For GitLab, prefer the pipeline UI/API or `glab ci` when available. For Azure DevOps, prefer injected failure evidence and the build/pipeline UI or REST when authenticated access is already available. For Bitbucket, prefer injected failure evidence and Bitbucket Pipelines UI/API when authenticated access is already available. For Gitea, prefer injected failure evidence and Gitea Actions UI/API when authenticated access is already available. Do not invent new CLI surfaces.</overview> <phase name="triage"> <steps> <step>Start from `triggering_run` when present; otherwise use only the latest failed default-branch run for the repository provider. Ignore older history.</step> <step>No-op with evidence if a newer same-workflow run already passes, the failure is a one-off flake, or an open Roomote PR already covers it.</step> </steps> </phase> <phase name="fix"> <steps> <step>Reproduce the failing job commands in this environment. If it does not reproduce, no-op with evidence.</step> <step>Implement the smallest fix, re-verify, and open a draft PR in this task.</step> <step>Resolve any existing investigating manager thread on closeout. Stay silent while working.</step> </steps> </phase></workflow><completion_criteria><criterion>Only the latest/trigger failure was considered.</criterion><criterion>Real failures were fixed and PR'd here, or no-opped with evidence when not fixable.</criterion></completion_criteria>
roocodeinc/roomote/tree/main/packages/cloud-agents/src/server/workflows/skills/standard/ci-failure-triage commit 76b0aa2f26
Frequently asked questions
Run npx skillmds@latest add roocodeinc/ci-failure-triage in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
Investigate the latest default-branch CI failure in an environment-backed workspace and, when it is real and fixable, fix and open a PR in the same task. It is listed under DevOps & Infra on SkillMD.
This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
RooCodeInc (@roocodeinc) published this skill. Their other Agent Skills are listed on their SkillMD profile.