playbook-native

Execute .playbook.md files as GitHub Action steps with a standalone Node runtime. No dependency on anthropics/claude-code-action; calls the Anthropic SDK directly, step by step.

Pairs with PLAYBOOK-MD/playbook-run (composite variant). Inputs and outputs are identical; swap uses: to switch runtimes.

When to choose playbook-native

• You want pinned wire behavior and deterministic post-step job-summary logs.
• Your playbook does not use MCP-dependent directives (@tool(mcp:...), @prompt(mcp:...), @prompt(library:...)). If it does, use playbook-run instead.
• You can't or don't want to use anthropics/claude-code-action.

Quick start

name: Run playbook
on: [workflow_dispatch]
jobs:
  run:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v5
      - uses: PLAYBOOK-MD/playbook-native@v1
        with:
          playbook: playbooks/greet.playbook.md
          inputs: |
            greeting: Hello
        env:
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}

Inputs

Input	Required	Default	Description
playbook	yes	—	Path to the .playbook.md file
inputs	no	``	YAML block of input values
elicit	no	``	YAML block of @elicit overrides keyed by step number
output-path	no	out/<run-id>.md	Destination for the final artifact
model	no	claude-opus-4-7	Anthropic model ID
strict	no	false	Fail on validation warnings

Outputs

Output	Description
artifact-path	Path to the written artifact
status	ok, failed, or skipped
step-count	Number of steps executed
warnings	Number of warnings emitted

Capability split vs playbook-run

See the ecosystem guide for the full compatibility matrix. Short version: native supports everything except MCP directives.

Examples

See examples/ for PR-review, scheduled, and workflow_dispatch workflows.