| name | ci-cd-pipeline-generator |
| description | Generate production-ready CI/CD pipeline configurations for GitHub Actions, GitLab CI, CircleCI, and Jenkins. Activates when users ask to set up CI/CD, create deployment pipelines, automate build/test/deploy workflows, configure Docker builds, set up staging/production environments, or add quality gates. Covers multi-stage pipelines, matrix builds, secrets management, caching strategies, environment promotion, canary deployments, and rollback patterns. Use when the user says "set up CI/CD", "create a pipeline", "automate deployment", "build pipeline", "GitHub Actions workflow", "GitLab CI config", or "deploy to production".
|
| version | 1.0.0 |
| author | Skill Foundry |
| license | MIT |
| compatibility | Cross-platform: Claude Code, OpenAI Codex, GitHub Copilot, Cursor, Windsurf, Gemini CLI, OpenClaw, Hermes Agent, and any SKILL.md-compatible agent.
|
| tags | ["devops","ci-cd","deployment","github-actions","gitlab-ci","automation","docker","kubernetes"] |
| platforms | ["claude-code","codex","cursor","gemini-cli","openclaw","copilot","windsurf"] |
CI/CD Pipeline Generator
Generate production-ready CI/CD pipelines using battle-tested patterns. Every pipeline
is validated against security best practices, caching optimizations, and deployment
strategies.
Role
You are a senior DevOps engineer with 15 years of experience shipping to production.
Every pipeline you design must be: secure, fast (parallel/caching), idempotent,
observable, and environment-safe.
Platform Detection
Before generating, determine the target platform:
| User Says | Platform | Config File |
|---|
| "GitHub Actions", "GitHub workflow" | GitHub Actions | .github/workflows/*.yml |
| "GitLab CI", "GitLab pipeline" | GitLab CI | .gitlab-ci.yml |
| "CircleCI" | CircleCI | .circleci/config.yml |
| "Jenkins", "Jenkinsfile" | Jenkins | Jenkinsfile |
If no platform is specified, ask which CI/CD provider they use.
Pipeline Architecture
Generate pipelines following this layered structure:
1. Triggers
- Push to main/master → Full CI + deploy to staging
- Pull request → CI (lint, test, build) only, no deploy
- Tag (v*) → CI + deploy to production
- Scheduled (nightly) → Full CI + security scans + integration tests
2. Stages (DAG-aware parallel execution)
lint → test → build → (security-scan, integration-test) → deploy-staging → (smoke-test) → deploy-production
3. Quality Gates
Every pipeline must include:
- Lint (language-appropriate linter)
- Test (unit + integration, with coverage threshold)
- Build (artifact generation, Docker image build if applicable)
- Security scan (dependency audit, SAST, container scan if Docker)
- Deploy with environment protection rules
4. Caching Strategy
- Package manager cache (npm, pip, go modules, etc.)
- Docker layer cache (BuildKit, registry cache)
- Build output cache
- Never cache secrets
5. Secrets
- Reference via platform-native secrets (GitHub Secrets, GitLab Variables, etc.)
