CI/CD

Overview

This skill covers the complete lifecycle of CI/CD pipeline design, implementation, and optimization across platforms including GitHub Actions, GitLab CI, Jenkins, CircleCI, and cloud-native solutions. It encompasses automated testing integration, security scanning, artifact management, deployment strategies, and specialized pipelines for ML workloads.

When to Use

• Implementing or migrating CI/CD pipelines
• Optimizing build and test execution times
• Integrating security scanning (SAST, DAST, dependency checks)
• Setting up deployment automation with rollback strategies
• Configuring test suites in CI environments
• Managing artifacts and container registries
• Implementing ML model training and deployment pipelines
• Troubleshooting pipeline failures and flakiness

Instructions

1. Analyze Requirements

• Identify build and test requirements
• Determine deployment targets and environments
• Assess security scanning needs (SAST, DAST, secrets, dependencies)
• Plan environment promotion strategy (dev → staging → production)
• Define quality gates and approval workflows
• Identify test suite composition (unit, integration, E2E)
• Determine artifact storage and retention policies

2. Design Pipeline Architecture

• Structure stages logically with clear dependencies
• Optimize for speed through parallelization and caching
• Design fail-fast strategy (lint → unit tests → integration tests → build)
• Plan secret management and secure credential handling
• Define deployment strategies (rolling, blue-green, canary)
• Architect for rollback and recovery procedures
• Design matrix builds for multi-platform support
• Plan monorepo CI strategies if applicable

3. Implement Testing Integration

• Configure unit test execution with coverage reporting
• Set up integration tests with service dependencies (databases, APIs)
• Implement E2E/smoke tests for critical user journeys
• Configure test parallelization and sharding
• Integrate test result reporting (JUnit, TAP, JSON)
• Set up flaky test detection and quarantine
• Configure performance/load testing stages
• Implement visual regression testing if applicable

4. Implement Security Scanning

• Integrate SAST (static analysis) tools (SonarQube, CodeQL, Semgrep)
• Configure DAST (dynamic analysis) for deployed environments
• Set up dependency/vulnerability scanning (Dependabot, Snyk, Trivy)
• Implement container image scanning
• Configure secrets detection (GitGuardian, TruffleHog)
• Set up license compliance checking
• Define security gate thresholds and failure policies

5. Implement Build and Artifact Management

• Configure dependency caching strategies
• Implement build output caching and layer caching (Docker)
• Set up artifact versioning and tagging
• Configure container registry integration
• Implement multi-stage builds for optimization
• Set up artifact signing and attestation
• Configure artifact retention and cleanup policies

6. Implement Deployment Automation

• Configure environment-specific deployments
• Implement deployment strategies (rolling, blue-green, canary)
• Set up health checks and readiness probes
• Configure smoke tests post-deployment
• Implement automated rollback on failure
• Set up deployment notifications (Slack, email, PagerDuty)
• Configure manual approval gates for production

7. Optimize Pipeline Performance

• Analyze pipeline execution times and bottlenecks
• Implement job parallelization for independent tasks
• Configure aggressive caching (dependencies, build outputs, Docker layers)
• Optimize test execution (parallel runners, test sharding)
• Use matrix builds efficiently
• Consider self-hosted runners for performance-critical workloads
• Implement conditional job execution (path filters, change detection)

8. Ensure Reliability and Observability

• Add retry logic for transient failures
• Implement comprehensive error handling
• Configure alerts for pipeline failures
• Set up metrics and dashboards for pipeline health
• Document runbooks and troubleshooting procedures
• Implement audit logging for deployments
• Configure SLO tracking for pipeline performance

Best Practices

Core Principles

1. Fail Fast: Run cheap, fast checks first (lint, type check, unit tests)
2. Parallelize Aggressively: Run independent jobs concurrently
3. Cache Everything: Dependencies, build outputs, Docker layers
4. Secure by Default: Secrets in vaults, least privilege, audit logs
5. Environment Parity: Keep dev/staging/prod as similar as possible
6. Immutable Artifacts: Build once, promote everywhere
7. Automated Rollback: Every deployment must be reversible
8. Idempotent Operations: Pipelines should be safely re-runnable

Testing in CI/CD

1. Test Pyramid: More unit tests, fewer integration tests, minimal E2E
2. Isolation: Tests should not depend on execution order
3. Determinism: Eliminate flaky tests or quarantine them
4. **Fast Feedba

...