✅ fresh

last synced 2026-06-17T22:39:35.514075+00:00 · coverage 100% (ai-techwriter-setup)

Validation by Beadloom doc_sync — same source as sync-check.

AI Tech-Writer Setup

The setup-ai-techwriter scaffolder, in the onboarding domain.

Source: src/beadloom/onboarding/ai_techwriter_setup.py

---

Specification

Purpose

Make adopting the AI tech-writer a one-command affair. beadloom setup-ai-techwriter --platform github|gitlab emits the CI workflow, the operator recipe, a provisioner, and the guide that wire the packaged harness (beadloom.ai_agents.ai_techwriter) into a target repo. The harness ships inside the installed beadloom package, so there is no Python vendoring — adopters depend on beadloom, and the scaffold only emits the workflow that invokes the installed module as python -m beadloom.ai_agents.ai_techwriter.

Output

scaffold(target_root, *, platform) writes the platform CI wrapper (a GitHub Actions job or a GitLab CI include), the operator recipe, the provisioner runner, and the guide. An unknown platform raises ValueError. Re-running cleanly overwrites the generated files.

Invariants

• No harness source is copied into the target repo; the workflow invokes the installed beadloom module.
• The scaffold is idempotent — generated files are cleanly overwritten on re-run.
• Only github and gitlab are accepted platforms.

API

Module src/beadloom/onboarding/ai_techwriter_setup.py:

• scaffold(target_root: Path, *, platform: str) -> list[Path] — emit the CI wrapper + recipe + provisioner + guide; returns the paths written.

Testing

Tests: tests/test_cli_setup_ai_techwriter.py, tests/test_ai_techwriter_cli.py