Skip to content

AI Skills

The GitHub Actions lexicon ships AI skills that teach AI coding agents (like Claude Code) how to build, validate, and deploy GitHub Actions workflows from a chant project.

Skills are structured markdown documents bundled with a lexicon. When an AI agent works in a chant project, it discovers and loads relevant skills automatically — giving it operational knowledge about the deployment workflow without requiring the user to explain each step.

When you scaffold a new project with chant init --lexicon github, skills are installed to .claude/skills/ for automatic discovery by Claude Code.

For existing projects, create the files manually:

.claude/
skills/
chant-github/
SKILL.md # skill content

The inline chant-github skill covers the full workflow lifecycle:

  • Buildchant build src/ --output .github/workflows/ci.yml
  • Validatechant lint src/
  • Deploy — commit and push the generated YAML
  • Status — GitHub Actions UI or gh run list
  • Troubleshooting — lint rule codes (GHA001–GHA020), post-synth checks (GHA006–GHA019)

The skill is invocable as a slash command: /chant-github

A file-based skill loaded from src/skills/github-actions-patterns.md. It provides pattern knowledge for:

  • Workflow structure — name, on, permissions, jobs
  • Trigger patterns — push, pull_request, schedule, workflow_dispatch
  • Matrix strategy — multi-OS, multi-version builds
  • Caching — SetupNode cache option, CacheAction composite
  • Permissions — least-privilege patterns
  • Reusable workflows — ReusableWorkflowCallJob
  • Artifacts — upload/download between jobs
  • Concurrency — group + cancel-in-progress

The lexicon provides MCP (Model Context Protocol) tools and resources that AI agents can use programmatically:

MCP toolDescription
diffCompare current build output against previous output
MCP resourceDescription
resource-catalogJSON list of all supported GitHub Actions entity types
examples/basic-ciExample CI workflow with TypeScript source