github-actions

name	github-actions
description	Comprehensive guide for building robust, secure, and efficient CI/CD pipelines using GitHub Actions. Use when the task involves `creating or editing GitHub Actions workflows`, `.github/workflows/*.yml`, `CI/CD pipelines`, `GitHub Actions best practices`, or `workflow optimization`.
license	MIT
metadata	{"version":"1.0.0"}

When to Use

• Creating or editing GitHub Actions workflow files (.github/workflows/*.yml)
• Reviewing CI/CD pipelines for security, performance, or correctness
• Setting up automated testing, building, or deployment pipelines
• Troubleshooting failing or flaky GitHub Actions workflows
• Designing deployment strategies (staging, production, blue/green, canary)

Critical Patterns

• Pin Actions to Commit SHA: Always use full commit SHA, never mutable tags (@v4, @main). Add version as comment: uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1. Tags can be silently moved to compromised commits (supply chain attack).
• Least Privilege GITHUB_TOKEN: Set permissions: contents: read at workflow level. Grant write only where explicitly needed, per-job.
• Secrets via secrets Context Only: Never hardcode sensitive data. Use ${{ secrets.NAME }}. Use environment secrets for deployment targets with manual approvals.
• OIDC Over Static Credentials: Use OpenID Connect for cloud auth (AWS, Azure, GCP) instead of long-lived access keys.
• Cache Dependencies: Use actions/cache with hashFiles() keys for node_modules, pip, Maven, etc. to dramatically speed up builds.
• Shallow Clone: Use fetch-depth: 1 in actions/checkout unless full history is needed.
• Fail Fast on Security: Integrate dependency review and SAST (CodeQL) as blocking checks.
• Environment Protection: Use GitHub Environments with required reviewers and branch restrictions for staging/production deploys.

Workflow Structure

name: CI/CD Pipeline

on:
  push:
    branches: [ main ]
  pull_request:
    branches: [ main ]
  workflow_dispatch:
    inputs:
      environment:
        description: 'Deploy target'
        required: false
        default: 'staging'
        type: choice
        options: [ staging, production ]

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

permissions:
  contents: read  # Least privilege default

jobs:
  lint:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
        with:
          fetch-depth: 1
      - uses: actions/setup-node@3235b876344d2a9aa001b8d1453c930bba69e610 # v3.9.1
        
           
           
         
          

  
     
     
    
         
         
        
           
           
         
           
         
         
        
           
           
           

  
     
     
    
         
    
         
         
        
           
           
         
          
         
         
                  
        
           
           
           

  
     
     
       
    
       
       
    
         
        
           
          
        
             
         

Code Examples

Action Pinning (Security)

# BAD: mutable tag — vulnerable to supply chain attacks
- uses: actions/checkout@v4
- uses: some-org/some-action@main

# GOOD: immutable SHA with version comment
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
- uses: some-org/some-action@a1b2c3d4e5f6a1b2c3d4e5f6a1b2c3d4e5f6a1b2 # v2.1.0

Permissions (Least Privilege)

# Workflow-level: restrictive default
permissions:
  contents: read

jobs:
  lint:
    # Inherits read-only — no override needed
    steps: [ ... ]

  deploy:
    permissions:
      contents: read
      deployments: write  # Only this job needs write
    steps: [ ... ]

  comment-on-pr:
    permissions:
      contents: read
      pull-requests: write  # Only this job needs PR write
    steps: [ ... ]

Caching

- name: Cache node modules
  uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
  with:
    path: |
      ~/.npm
      node_modules
    key: ${{ runner.os }}-node-${{ hashFiles('**/package-lock.json') }}
    restore-keys: |
      ${{ runner.os }}-node-

Matrix Strategy

jobs:
  test:
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ ubuntu-latest, windows-latest ]
        node-version: [ 18, 20, 22 ]
        exclude:
          - os: windows-latest
            node-version: 18
    steps:
      - uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
      - uses: actions/setup-node@3235b876344d2a9aa001b8d1453c930bba69e610 # v3.9.1
        with:
          node-version: ${{ matrix.node-version }}
      - run: npm ci && npm test

OIDC Cloud Authentication

jobs:
  deploy:
    runs-on: ubuntu-latest
    permissions:
      id-token: write   # Required for OIDC
      contents: read
    steps:
      - uses: aws-actions/configure-aws-credentials@ececac1a45f3b08a01d2dd070d28d111c5fe6722 # v4.1.0
        with:
          role-to-assume: arn:aws:iam::123456789012:role/my-deploy-role
          aws-region: us-east-1
      # No static credentials stored — short-lived token via OIDC
      - run: aws s3 sync dist/ s3://my-bucket/

Integration Tests with Services

jobs:
  integration:
    runs-on: ubuntu-latest
    services:
      postgres:
        image: postgres:16-alpine
        env:
          POSTGRES_DB: testdb
          POSTGRES_USER: test
          POSTGRES_PASSWORD: test
        ports:
          - 5432:5432
        options: >-
          --health-cmd "pg_isready -U test"
          --health-interval 10s
          --health-timeout 5s
          --health-retries 5
      redis:
        image: redis:7-alpine
        ports:
          - 6379:6379
    steps:
      - uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
      - run: npm ci
      - run: npm run test:integration
        env:
          DATABASE_URL: postgresql://test:test@localhost:5432/testdb
          REDIS_URL: redis://localhost:6379

Environment Protection (Production Deploy)

jobs:
  deploy-prod:
    runs-on: ubuntu-latest
    needs: [ test, build, deploy-staging ]
    if: github.ref == 'refs/heads/main'
    environment:
      name: production
      url: https://example.com
    steps:
      - uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
        with:
          name: app-build
      - name: Deploy to production
        env:
          PROD_API_KEY: ${{ secrets.PROD_API_KEY }}
        run: ./deploy.sh --env production
      - name: Post-deploy smoke test
        run: |
          curl -sf https://example.com/health || exit 1
          echo "Smoke test passed"

Reusable Workflow

# .github/workflows/reusable-build.yml
name: Reusable Build
on:
  workflow_call:
    inputs:
      node-version:
        required: false
        type: string
        default: '20'
    outputs:
      artifact-name:
        description: 'Name of the uploaded artifact'
        value: ${{ jobs.build.outputs.artifact-name }}

jobs:
  build:
    runs-on: ubuntu-latest
    outputs:
      artifact-name: app-build-${{ github.sha }}
    steps:
      - uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
      - uses: actions/setup-node@3235b876344d2a9aa001b8d1453c930bba69e610 # v3.9.1
        with:
          node-version: ${{ inputs.node-version }}
          cache: 'npm'
      - run: npm ci && npm  
         
        
             
           




  
     
    
       

Anti-Patterns to Avoid

Anti-Pattern	Risk	Do Instead
uses: actions/checkout@v4 (mutable tag)	Supply chain attack — tag can be moved to malicious commit	Pin to full SHA: @34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
No permissions block	GITHUB_TOKEN gets default broad access	Set permissions: contents: read at workflow level
secrets.MY_KEY in echo or logs	Secret leakage even with masking	Never print secrets; use them only in env vars for commands
Long-lived cloud credentials as secrets	Large blast radius if compromised	Use OIDC for short-lived tokens
fetch-depth: 0 by default	Slow checkout, wastes bandwidth	fetch-depth: 1 unless full history needed
No concurrency on deploy workflows	Race conditions, double deploys	Add concurrency with cancel-in-progress
Hardcoded versions in multiple places	Drift, maintenance burden	Use reusable workflows or composite actions
No timeout-minutes on jobs	Hung workflows burn runner minutes	Set reasonable timeout-minutes per job
No retention-days on artifacts	Storage bloat, cost increase	Set retention-days based on need (7-30 days)
Running E2E tests on every push	Slow CI, wasted resources	Run on PR and main only, or use path filters

Deployment Strategies Reference

Strategy	How It Works	Best For	Rollback
Rolling	Gradually replaces old instances	Stateless apps, most cases	Redeploy previous version
Blue/Green	Full parallel env, switch traffic	Zero-downtime critical apps	Switch traffic back to blue
Canary	Route small % to new version	Risk-sensitive changes	Route 100% back to stable
Dark Launch	Deploy hidden behind feature flags	Decoupling deploy from release	Toggle flag off

Commands

# Validate workflow syntax locally
actionlint .github/workflows/*.yml

# Run workflow locally with act
act push --job build

# List workflow runs
gh run list --workflow=ci.yml

# View specific run logs
gh run view <run-id> --log

# Re-run failed jobs
gh run rerun <run-id> --failed

# Trigger manual workflow
gh workflow run deploy.yml -f environment=staging

# List secrets (names only)
gh secret list

Resources

• Review Checklist: See references/review-checklist.md for a comprehensive workflow review checklist
• Troubleshooting: See references/troubleshooting.md for diagnosing common failures
• GitHub Actions Documentation
• Security Hardening Guide
• OIDC with Cloud Providers
• Related skill: docker-expert — for Dockerfile and container best practices