How do I install the CI Triage skill?

Run npx skillmds@latest add mentra-community/ci-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.

Is CI Triage safe to use?

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.

Which AI agents work with CI Triage?

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.

Is CI Triage free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published CI Triage?

mentra-community (@mentra-community) published this skill. Their other Agent Skills are listed on their SkillMD profile.

CI Triage by mentra-community — Agent Skill

Name: CI Triage
Author: mentra-community

CI Triage

Triage failing GitHub PR checks: list failures with gh, fetch capped Actions logs, skip non-Actions checks, and summarize root cause. Use when the user pastes CI failures, asks why a PR is red, mentions failing checks, or says /ci.

File contents

CI failure triage

Never paste multi-thousand-line logs into chat

Summarize root causes. Cap every log fetch.

Protocol

1. Resolve PR number

Explicit arg from the user (e.g. /ci 523 or "PR 3502"), or
gh pr view --json number -q .number on the current branch.

2. List failing checks

gh pr checks <n> --json name,link,workflow,bucket,state

Filter to entries where bucket == "fail".

Field name is workflow, not workflowName.

3. Classify each failure

For each failing check, test link against:

/\/actions\/runs\/(\d+)/

Match	Meaning	Next step
Yes	GitHub Actions run	Extract `runId`, go to step 4
No	External check (Cloudflare Pages, Vercel, Bugbot, etc.)	Do not call `gh run view`. Carry forward as name + link + state only

4. Dedupe Actions runs

Multiple failing jobs can share one runId. Deduplicate by runId before fetching logs.

5. Fetch bounded logs (max 3 runs)

For each unique runId, stop after the first 3 distinct runs. Extra runs: list by check name only, with a note that more failures exist.

gh run view <runId> --log-failed | tail -c 20000

Always pipe through tail -c 20000 (~20KB). Never feed the unbounded stream into context.

6. Root-cause analysis

Prefer the built-in Task subagent when available:

Task(
  subagent_type: "ci-investigator",
  description: "Investigate PR <n> check <name>",
  prompt: "<PR number, repo, run id, failing check name, and the capped log excerpt>"
)

If ci-investigator is not available in this agent runtime, analyze the capped log excerpt yourself (or use any local agent definition under ~/.cursor/agents/ / .cursor/agents/ if present). Do not stall the triage on a missing subagent type.

7. Reply format

One aggregated report:

Actions root causes — short summaries (per run/check).
Non-Actions failures — name, state, link only (open in browser).
No raw log pastes.

If nothing failed, say so and stop.

---name: ci-triagedescription: Triage failing GitHub PR checks: list failures with gh, fetch capped Actions logs, skip non-Actions checks, and summarize root cause. Use when the user pastes CI failures, asks why a PR is red, mentions failing checks, or says /ci.---# CI failure triage## Never paste multi-thousand-line logs into chatSummarize root causes. Cap every log fetch.## Protocol### 1. Resolve PR number- Explicit arg from the user (e.g. `/ci 523` or "PR 3502"), **or**- `gh pr view --json number -q .number` on the current branch.### 2. List failing checks```bashgh pr checks <n> --json name,link,workflow,bucket,state```Filter to entries where `bucket == "fail"`.**Field name is `workflow`, not `workflowName`.**### 3. Classify each failureFor each failing check, test `link` against:```/\/actions\/runs\/(\d+)/```| Match | Meaning | Next step ||---|---|---|| Yes | GitHub Actions run | Extract `runId`, go to step 4 || No | External check (Cloudflare Pages, Vercel, Bugbot, etc.) | **Do not** call `gh run view`. Carry forward as name + link + state only |### 4. Dedupe Actions runsMultiple failing jobs can share one `runId`. Deduplicate by `runId` before fetching logs.### 5. Fetch bounded logs (max 3 runs)For each unique `runId`, **stop after the first 3 distinct runs**. Extra runs: list by check name only, with a note that more failures exist.```bashgh run view <runId> --log-failed | tail -c 20000```Always pipe through `tail -c 20000` (~20KB). Never feed the unbounded stream into context.### 6. Root-cause analysisPrefer the built-in Task subagent when available:```Task( subagent_type: "ci-investigator", description: "Investigate PR <n> check <name>", prompt: "<PR number, repo, run id, failing check name, and the capped log excerpt>")```If `ci-investigator` is not available in this agent runtime, analyze the capped log excerpt yourself (or use any local agent definition under `~/.cursor/agents/` / `.cursor/agents/` if present). Do not stall the triage on a missing subagent type.### 7. Reply formatOne aggregated report:1. **Actions root causes** — short summaries (per run/check).2. **Non-Actions failures** — name, state, link only (open in browser).3. No raw log pastes.If nothing failed, say so and stop.

Frequently asked questions