cicd_skill
- Shell
2
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 pluginagentmarketplace/custom-plugin-devops --skill cicd- SKILL.md2.5 KB
Overview
This skill helps teams design, implement, and troubleshoot CI/CD pipelines across Git, GitHub Actions, GitLab CI, and Jenkins. It focuses on deployment strategies, Git workflows, and practical automation for reliable delivery. The goal is to reduce manual steps, improve deployment safety, and accelerate feedback loops.
How this skill works
The skill inspects repository structure, pipeline definitions, and common pipeline parameters to recommend improvements and fixes. It provides concrete commands, workflow templates, and rollback procedures, and explains when to apply rolling, blue-green, or canary deployments. It also surfaces troubleshooting steps for common failures like timeouts, flaky tests, and missing secrets.
When to use it
- Setting up or migrating CI/CD pipelines for a new or existing repository
- Standardizing Git workflows and pull request practices across teams
- Implementing safe deployment strategies (rolling, blue-green, canary)
- Diagnosing pipeline failures, flaky tests, and secret/config issues
- Automating promotion between environments and rollback procedures
Best practices
- Adopt a simple Git workflow (trunk-based or GitFlow) and enforce PR checks
- Keep pipeline steps modular with caching to reduce build time
- Store secrets in platform-managed secret stores, not in code
- Use environment promotion and feature flags to reduce blast radius
- Add clear rollback commands and validate them in runbooks
Example use cases
- Create a GitHub Actions workflow that builds, tests, and deploys to staging on PR merge
- Migrate Jenkins jobs to GitLab CI with pipeline-as-code and shared templates
- Implement blue-green deployment for zero-downtime releases to production
- Add caching and parallel test execution to cut CI time and reduce timeouts
- Run a debug checklist to resolve a failing pipeline: syntax, logs, secrets, local run
FAQ
Choose rolling for low-risk incremental updates, blue-green for zero-downtime releases, and canary when you need staged exposure and fast rollback.
How do I recover from a failed deployment?
Run the rollback command (for Kubernetes: kubectl rollout undo), collect deployment and pipeline logs, fix the root cause, then redeploy to a validated environment.