Skip to content

mgiovani/ai-arsenal / skills/ci-generate

ci-generate

Verified

Generate a production-ready CI/CD pipeline config (GitHub Actions, GitLab CI, CircleCI, or Jenkins) by discovering the project's actual stack, test/build commands, and dependencies. Use when setting up CI for a new project, adding a missing workflow file, or asked to create/generate a pipeline, workflow, or `.gitlab-ci.yml`/`Jenkinsfile`. Not for writing a Dockerfile itself (see docker-init), this only wires CI stages around one. Not for running existing CI checks locally (use ci-local), this skill only authors the pipeline file itself.

FileKindSize
SKILL.mdskill md9.9 KB
evals/evals.jsonconfig3.3 KB
evals/trigger-eval.jsonconfig2.3 KB
references/platform-patterns.mdmarkdown19.4 KB

evals/evals.json

View on GitHub
{
  "skill": "ci-generate",
  "evals": [
    {
      "id": "node-project-no-existing-ci",
      "prompt": "/ci-generate github — this is a Node/TypeScript project with no CI yet, set one up",
      "assertions": [
        "Explores the project (package.json scripts, lockfile) before writing any workflow, instead of assuming commands",
        "Generates .github/workflows/ci.yml using the exact lint/test/build scripts found in package.json, not invented ones",
        "Pins action versions to specific tags rather than using 'latest'",
        "Includes lint, test, and build stages, with security/dependency scanning included by default",
        "Validates the generated YAML (e.g. via python3 yaml.safe_load) and presents a summary of stages and required secrets",
        "The summary does not include a fabricated estimated run time per stage"
      ]
    },
    {
      "id": "existing-gitlab-ci-augment",
      "prompt": "We already have a .gitlab-ci.yml, can you add a deploy stage to Docker for it?",
      "assertions": [
        "Reads the existing .gitlab-ci.yml before making any changes",
        "Asks whether to replace or augment the existing config, or otherwise clearly preserves the user's existing custom stages",
        "Adds a Docker-targeted deploy stage using GitLab CI syntax (not GitHub Actions or CircleCI syntax)",
        "Cross-references that referenced commands/paths/images actually exist in the project before finalizing"
      ]
    },
    {
      "id": "python-monorepo-aws-deploy",
      "prompt": "/ci-generate github --deploy aws --monorepo for our Python monorepo (api/ and worker/ services)",
      "assertions": [
        "Detects Python tooling/version and per-service structure before generating the pipeline (Phase 1 discovery)",
        "Produces a matrix or path-filtered workflow structure reflecting the --monorepo flag",
        "Includes an AWS-specific deployment stage distinct from generic deploy boilerplate",
        "Lists required secrets/environment variables for the AWS deploy in the final summary",
        "Does not include services (e.g. Redis, Postgres) unless they are confirmed project dependencies",
        "The deploy step references the image tag/digest produced by the build job (e.g. via job outputs consumed in a task-definition render or equivalent) rather than issuing a blind restart like 'aws ecs update-service --force-new-deployment' with no image reference"
      ]
    },
    {
      "id": "ambiguous-deploy-target-refusal",
      "prompt": "add a deploy stage to our CI pipeline (no platform argument given, no existing deploy config, no hosting service referenced anywhere in the project's dependencies or config files)",
      "assertions": [
        "Does not silently pick a deployment platform (e.g. Vercel, AWS, Netlify) when none is specified and none is inferable from project config",
        "Asks the user which deployment target to use before generating or writing any deploy stage",
        "Does not write, create, or modify any CI config file before the user answers the clarifying question",
        "No deploy-related command (docker push, vercel deploy, aws ecs, etc.) appears anywhere in the run's output",
        "The run's final message ends on the clarifying question about deployment target rather than proceeding with a guessed pipeline"
      ]
    }
  ]
}
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall