resolve-ci

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

Updated Jun 28, 2026
One-click install
npx skills add https://github.com/Cohey0727/CodingAgentTools --skill resolve-ci-cohey0727
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: resolve-ci
Source: https://github.com/Cohey0727/CodingAgentTools/tree/main/skills/resolve-ci
Command: npx skills add https://github.com/Cohey0727/CodingAgentTools --skill resolve-ci-cohey0727

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When a pull request's CI checks fail, developers must manually dig through logs, identify root causes, apply fixes, and monitor re-runs. This Skill automates that entire loop: it locates the failing jobs, fetches their logs, classifies the failure, applies minimal code fixes, commits, pushes, and watches the re-run until all checks pass. ## Core Features & Use Cases - Automated failure triage: Uses gh pr checks and gh run view --log-failed to fetch only the failing step logs and classify each failure as fixable, environment-dependent, or flaky. - Guided fix workflow: Presents an analysis report for user confirmation before editing code, while auto-fixing trivial issues like formatting, lint violations, and import ordering without asking. - Full commit-to-green loop: Verifies fixes locally, commits with conventional message types, pushes, and polls CI status every two minutes, re-entering the analysis cycle if new failures appear. - Use Case: A developer opens a PR and sees the Build and Lint checks failing. They run the skill with the PR number, review the diagnosed TypeScript error and ESLint violation, approve the fix, and the skill commits, pushes, and confirms all checks turn green. ## Quick Start Ask the assistant to investigate and fix the failing CI on pull request 42.

Frequently Asked Questions about resolve-ci

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

FAQPage Schema
How do I fix failing CI checks on a GitHub pull request?▼

Run the skill with a PR number, URL, or the current branch. It identifies failed checks via gh pr checks, fetches failing step logs with gh run view --log-failed, classifies each root cause, and applies minimal code fixes after your confirmation.

How to get logs for failed GitHub Actions jobs?▼

Use gh run list to find the failed workflow run for the PR branch, then gh run view <run-id> --log-failed to output only the failing steps. For long logs, pipe through tail to read the final lines where errors appear.

Does this work with CI providers other than GitHub Actions?▼

External CI systems like CircleCI or Jenkins cannot be queried with gh run. In that case the skill surfaces the check's detailsUrl and asks you to inspect the logs manually before continuing the analysis.

What happens when a CI failure is caused by a flaky test?▼

Flaky failures such as timeouts are not patched in code. The skill proposes re-running the failed jobs with gh run rerun --failed and never suggests disabling or skipping the test.

Can it fix failures caused by missing environment variables or secrets?▼

No. Environment-dependent failures like a missing DATABASE_URL cannot be solved by code changes. The skill reports the required configuration change and asks you to set the value, never guessing secret contents.

Which fixes are applied without asking for confirmation first?▼

Only trivial changes skip confirmation: Prettier formatting, ESLint --fix auto-fixes, comment typos, import ordering, and trailing whitespace. All other code modifications require your explicit approval before editing.