ci-triage

Classify failed GitHub Actions CI runs into failure categories and write handoff packets.

37|2|Updated Jun 28, 2025
One-click install
npx skills add https://github.com/jellyrock/jellyrock --skill ci-triage-jellyrock
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ci-triage
Source: https://github.com/jellyrock/jellyrock/tree/main/.claude/skills/ci-triage
Command: npx skills add https://github.com/jellyrock/jellyrock --skill ci-triage-jellyrock

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

When a GitHub Actions workflow fails on a PR or main branch, developers waste time scrolling through logs to find the root cause. This Skill automates the triage process, classifying failures into known categories and preparing a structured investigation handoff so you can fix issues faster.

Core Features & Use Cases

  • Automated CI Triage: Fetches failed GitHub Actions runs, extracts failure tails, and classifies them into categories like lint-fail, build-fail, device-test-fail, docs-stale-blocking, or language-coverage-fail.
  • Dedup-First Workflow: Checks for prior triage on the same run-id and short-circuits if nothing has changed, avoiding redundant work.
  • Structured Handoff: Writes a detailed handoff packet with file context and failure logs to the .claude/handoffs/ directory for cross-session resume and compaction recovery.
  • Use Case: A PR breaks the build due to a BSC type mismatch. Use this Skill to get an instant classification, the failing file, and a ready-to-use investigation contract.

Quick Start

Use the ci-triage skill to investigate the failing GitHub Actions run with ID 1234567890 and tell me what went wrong.

Frequently Asked Questions about ci-triage

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automatically triage failing GitHub Actions CI workflows?

You can automate CI failure triage by extracting run logs with the gh CLI and analyzing git diffs to classify failures into categories like lint-fail, build-fail, or device-test-fail. This process prepares a structured investigation handoff with file context to speed up root cause identification.

What is the best way to debug a continuous integration build failure?

The best way to debug a continuous integration build failure is to automate the extraction of run logs and git diffs to classify the error type. This avoids manual log scrolling and generates a structured handoff packet with the specific failing files for targeted investigation.

Do I need the gh CLI installed to investigate CI workflow failures?

Yes, the gh CLI is required to extract GitHub Actions run logs for CI triage. The triage process also uses git for diff analysis to classify failures and assemble the initial file context needed for investigation.

Can I use automated triage to deduplicate GitHub Actions failure analysis?

Yes, automated CI triage supports deduplicating GitHub Actions failure analysis. It checks for prior triage on the same run-id and short-circuits if nothing has changed, ensuring that continuous integration failures are only classified once to avoid redundant work.

How does CI triage handle cross-session resume after compaction recovery?

CI triage handles cross-session resume by writing structured handoff packets with failure logs and file context to the .claude/handoffs/ directory. This allows investigations into CI failures to be paused and resumed across different sessions or after compaction recovery.

What categories does automated CI triage use to classify GitHub Actions failures?

Automated CI triage classifies GitHub Actions failures into lint-fail, build-fail, device-test-fail, docs-stale-blocking, and language-coverage-fail categories. This classification helps pinpoint the exact nature of the continuous integration workflow failure for faster resolution.