fix-ci-failures

Diagnose and fix failing CI checks on GitHub pull requests using the gh CLI.

1|Updated Aug 27, 2026
One-click install
npx skills add https://github.com/Niiihuel/openide --skill fix-ci-failures-niiihuel
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fix-ci-failures
Source: https://github.com/Niiihuel/openide/tree/main/vscode/.github/skills/fix-ci-failures
Command: npx skills add https://github.com/Niiihuel/openide --skill fix-ci-failures-niiihuel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When CI checks fail on a pull request, finding the actual cause means digging through thousands of lines of GitHub Actions logs, artifacts, and job output. This Skill provides a structured workflow to locate failed checks, download the right logs, extract the root cause, and determine whether the failure is caused by your code or by flaky infrastructure. ## Core Features & Use Cases - Failed Check Discovery: Identify the PR for the current branch, list failed checks, and extract run and job IDs from check URLs. - Log Retrieval Strategies: View failed step logs directly, download test log and crash dump artifacts, or fetch per-job logs via the GitHub API while a run is still in progress. - Failure Triage: Distinguish real code regressions from infrastructure issues like npm registry timeouts, Electron CDN failures, disk space exhaustion, and flaky tests, then re-run failed jobs when appropriate. - Use Case: A PR shows a failing "Linux / Electron" integration test. Use this Skill to download the logs-linux-x64-electron-1 artifact, inspect exthost.log for the assertion failure, fix the code, push, and watch the checks turn green. ## Quick Start Find the failing CI checks on my current pull request, download the logs, and help me fix the root cause.

Frequently Asked Questions about fix-ci-failures

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

FAQPage Schema
How do I find failed CI checks on a GitHub pull request?▼

Run gh pr checks --json name,state,link,bucket to list all checks, then filter with --jq '.[] | select(.bucket == "fail")' to show only failures. The link field contains the run ID and job ID needed to fetch logs.

How do I download GitHub Actions logs for a failed job?▼

Use gh run view <RUN_ID> --job <JOB_ID> --log-failed for failed step output once the run completes. If the run is still in progress, call gh api repos/<owner>/<repo>/actions/jobs/<JOB_ID>/logs and save the output to a file before searching.

How can I tell if a CI failure is caused by my PR or is flaky?▼

Check whether the failing test relates to code you changed, then compare against recent main branch runs with gh run list --branch main. Infrastructure signs like ETIMEDOUT, npm registry errors, or 'No space left on device' indicate the failure is not from your code.

Why does gh run view --log-failed hang or error?▼

The --log-failed flag requires the entire run to complete, not just the failed job. If other jobs are still running, download the individual job log through the GitHub API instead, which works while the run is in progress.

How do I re-run only the failed jobs in a GitHub Actions workflow?▼

Run gh run rerun <RUN_ID> --failed to re-execute only the jobs that failed. This is the recommended action when failures are caused by infrastructure issues or flaky tests rather than code changes.