Configure a GitLab CI/CD Pipeline with Multi-Stage Deployment
Create a production-grade GitLab CI/CD pipeline with linting, testing, security scanning, and multi-environment deployments.
📝 The Prompt
Generate a complete `.gitlab-ci.yml` configuration for a [PROJECT_TYPE] project (e.g., Python microservice, React SPA, Go CLI tool) that implements a multi-stage CI/CD pipeline.
**Project Context:**
- Language/Runtime: [LANGUAGE_AND_VERSION]
- Build tool: [BUILD_TOOL]
- Deployment targets: [DEPLOYMENT_TARGETS] (e.g., AWS ECS, Kubernetes, S3 + CloudFront)
- Branch strategy: `main` (production), `develop` (staging), feature branches (review)
**Pipeline Stages to Include:**
1. **🔍 Lint & Static Analysis**: Run [LINTER_TOOLS] with fail-on-warning. Cache linter dependencies.
2. **🧪 Test**: Execute unit and integration tests with coverage threshold of [COVERAGE_PERCENT]%. Generate JUnit XML reports and Cobertura coverage reports for GitLab merge request integration.
3. **🏗️ Build**: Compile/bundle the application. For containerized apps, build a Docker image using multi-stage Dockerfile, tag with `$CI_COMMIT_SHORT_SHA` and `latest`, and push to [CONTAINER_REGISTRY].
4. **🔒 Security Scan**: Integrate SAST, dependency scanning, and container scanning using GitLab templates or [SECURITY_TOOL].
5. **🚀 Deploy Staging**: Auto-deploy to staging on `develop` branch merges. Use [DEPLOYMENT_METHOD] with environment URL and rollback capability.
6. **🚀 Deploy Production**: Manual approval gate for production deployment from `main`. Include smoke test job that runs after deployment.
**Also include:**
- Global caching strategy for [DEPENDENCY_MANAGER] (e.g., npm, pip, go mod)
- `rules:` syntax (not `only/except`) for conditional job execution
- Environment-specific variables using GitLab CI/CD variable groups
- `needs:` for DAG pipeline optimization
- Retry configuration for flaky network-dependent jobs
- Clear inline comments explaining each decision
💡 Tips for Better Results
Specify your exact runner tags and executor type (Docker, Kubernetes, shell) since image availability and caching strategies differ significantly. Use the `needs` keyword to create a DAG pipeline — this can cut total pipeline time by 40-60% compared to sequential stages. Store secrets as masked and protected CI/CD variables and reference them by name in your prompt so the output uses proper variable substitution.
🎯 Use Cases
DevOps engineers and development teams use this when setting up or modernizing their GitLab CI/CD pipeline, especially during new project bootstrapping or migration from other CI systems like Jenkins or GitHub Actions.