deployment-pipeline-design_skill
- Shell
168
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill rsmdt/the-startup --skill deployment-pipeline-design- SKILL.md12.4 KB
Overview
This skill helps design CI/CD pipelines and deployment strategies, including blue-green, canary, and rolling releases. It covers pipeline architecture, quality gates, rollback patterns, and platform-specific patterns for GitHub Actions and GitLab CI. Use it to create reliable, observable, and secure automated release workflows.
How this skill works
The skill outlines a stage-based pipeline model (Build -> Test -> Analyze -> Package -> Deploy -> Verify) and maps deployment strategies to concrete steps and triggers. It provides patterns for artifact management, parallelization, caching, and environment parity, plus examples for GitHub Actions and GitLab CI. It also defines quality gates, manual approvals, rollback mechanisms, and security integration points.
When to use it
- Designing a new CI/CD pipeline from scratch for web or microservice apps
- Implementing blue-green, canary, or rolling deployment strategies
- Configuring quality gates, manual approvals, and environment protection
- Integrating SAST/DAST, dependency scanning, and secret management
- Creating automated rollback policies and migration-safe workflows
Best practices
- Fail fast: run lint and unit tests before expensive jobs
- Build immutable artifacts once and deploy the same artifact everywhere
- Use parallel jobs and caching to keep main branch pipelines under 15 minutes
- Enforce quality gates: unit/integration tests, coverage, and critical security checks
- Require multi-approver, time-boxed manual gates for production deployments
Example use cases
- Blue-green: deploy to standby environment, run smoke tests, switch load balancer for zero-downtime release
- Canary: shift 5% -> 25% -> 50% traffic with metric-based rollback triggers
- Rolling: replace instances in batches with maxUnavailable/maxSurge controls for constrained infra
- Feature flags: decouple deployment from release to enable safe rollouts and kill switches
- GitHub Actions reusable workflows: centralize deploy logic and call from multiple repos
FAQ
Choose canary when you need gradual validation with real traffic and limited infrastructure; choose blue-green for instant rollback and strict zero-downtime with duplicate environments.
How do I handle database schema changes safely?
Prefer forward-only migrations, make changes non-destructive, use feature flags for behavioral switches, and test rollback migrations in staging before production.
What triggers should cause an automated rollback?
Use metric-based triggers such as error rate exceeding baseline+threshold, p99 latency breaches, or repeated health-check failures within a short window.
