skills.rest
remotion-dev
remotion-devOFFICIAL
Explore all skills by this author →
Website𝕏@remotionGitHub

flake

Track and rerun flaky Remotion CI failures in GitHub Actions.

57.9k|4.4k|Updated Jun 23, 2020
Search Tags:#gh CLI#ci#Remotion#GitHub actions#workflow rerun#flake tracking
One-click install
npx skills add https://github.com/remotion-dev/remotion --skill flake
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: flake
Source: https://github.com/remotion-dev/remotion/tree/main/.agents/skills/flake
Command: npx skills add https://github.com/remotion-dev/remotion --skill flake

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Flaky CI checks waste developer time and obscure real regressions. This Skill centralizes flake tracking in issue #8375 and automates the rerun workflow so transient failures stop blocking pull requests.

Core Features & Use Cases

  • Failure Discovery: Inspect PRs, workflow runs, and recent failed jobs via the gh CLI when no PR context is provided.
  • Signature Tracking: Increment counts for repeated failure signatures or append new rows to the tracker table with workflow, job, first-seen date, and evidence URL.
  • Targeted Reruns: Rerun only failed jobs, cancel in-progress flaky suites, and verify the tracker after edits.
  • Use Case: A maintainer notices the Install and Test workflow failed on PR #4521 with a fetch failed error. Use this Skill to record the signature in issue #8375 and rerun the failed job to confirm it was a flake.

Quick Start

Use the flake skill to investigate the failed CI check on PR 4521, update the tracker issue, and rerun the flaky job.

Frequently Asked Questions about flake

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

FAQPage Schema
How do I rerun a failed GitHub Actions job for a pull request?▼

Use `gh run rerun <run-id> --repo remotion-dev/remotion --failed` to rerun only the failed jobs from a workflow run. For a single job, call `gh api --method POST repos/remotion-dev/remotion/actions/jobs/<job-id>/rerun` instead.

What counts as a flaky CI failure versus a real bug?▼

A failure is flaky if the same job passes on rerun without code changes, or if the error is infrastructure-related such as fetch failures, timeouts, port collisions, or runner provisioning errors. Deterministic type errors, lint failures, and assertion mismatches caused by source changes are not flakes.

How do I find failed CI checks when no PR number is given?▼

List recent open PRs with `gh pr list` and recent failed runs with `gh run list --workflow "Install and Test" --status failure`. Then inspect the run details and job logs to identify the failure signature.

Why does gh run watch exit non-zero after cancelling a workflow?▼

Cancellation marks the run as failed, so `gh run watch` returns a non-zero exit status. This is expected behavior and you should proceed to rerun the failed jobs after the cancellation is visible.

Can I update the flake tracker issue without write permissions on the GitHub app?▼

The Skill recommends using the `gh` CLI directly because the GitHub app may lack Actions log access or write permissions. Authenticate `gh` with an account that has write access to remotion-dev/remotion before editing issue #8375.

Related Skills

Explore More Software Engineering Skills→
DaddyElonMusk69DaddyElonMusk69

github-issues

Manage GitHub issues via gh CLI or REST API.

Community
Intermediate
AbsolutSystemsAbsolutSystems
1

debug

Trace software failures to their root cause before fixing.

Official
Advanced
iterateiterate
198

architect-diagnose-errors

Cluster Cloudflare and PostHog error signatures to identify root-cause groups.

Official
Intermediate
berkayturanciberkayturanci
4

keel-flake-audit

Detect intermittently failing tests and track them as issues.

Community
Advanced