review-ci

Your job is to check the CI status of the given Pull Request and explain whether checks are passing or failing, and why.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "review-ci" with this command: npx skills add saurabhdaware/agent-toolkit/saurabhdaware-agent-toolkit-review-ci

Your job is to check the CI status of the given Pull Request and explain whether checks are passing or failing, and why.

Arguments

  • PR_URL: GitHub Pull Request URL (preferred)

Use AskQuestion to get PR URL / PR Number / Repository if not provided in prompt.

Variables

Extract following variables from {{PR_URL}}. We'll use them ahead

  • REPO: '{{owner}}/{{reponame}}'

  • PR_NUMBER: {{number}}

E.g. in PR_URL: https://github.com/saurabhdaware/agent-toolkit/pull/123, REPO=saurabhdaware/agent-toolkit and PR_NUMBER=123

Steps to Review CI

Step 1: Getting status of all CI pipelines

Run the following command

gh pr checks {{PR_NUMBER}} --json name,state,description,link,event,bucket,workflow,startedAt,completedAt --repo {{REPO}}

  • If error in authentication or command installation, refer to gh installation and authentication guide

Step 2: Print a table with overall CI status

Print a table with following columns-

  • Workflow and Name ("workflow: name" format)

  • State (Use ✅, ❌, ⚠️ emojis for ease of reading)

  • Link

  • Description (if any)

Step 3: Deep dive on failures (if there are any failures)

Now we'll find the logs to see why CI checks are failing.

For each failing github actions CI checks (run it as separate subagents if possible),

  • Extract RUN_ID and JOB_ID from the Link

  • Command to get full logs gh run view {{RUN_ID}} --job {{JOB_ID}} --log --repo {{REPO}} (Logs can be large so smartly use grep to filter errors and their reasons)

  • You can also use {{PR_URL}}.diff to get diff of PR and see the changes in code that could be responsible for failures (diffs can be large too. Use grep to ignore snapshots, lock file changes, or similar irrelevant changes. Do not fetch diff again if it was already done in earlier chat)

Step 4: Print errors and their reasons (if there are any failures)

List down your findings on why each check is failing, what code change is responsible (if known), how to fix (if known)

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

review-ui

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

plan-and-execute

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

yolo

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

review-pr

No summary provided by upstream source.

Repository SourceNeeds Review