Agent Acceptance Test Harness Builder

WHY THIS EXISTS

Operational proof for AI-era sites and agents.

Convert an AI-agent task into acceptance tests, permission gates, stop rules, eval cases, Playwright-style smoke skeleton, and a JSON completion receipt. The useful output is a visible table plus a receipt that names input, checks, limits, and next action.

• Turns agent work into explicit objective, permissions, stop conditions, and proof artifacts.
• Generates happy-path, missing-input, permission, evidence-gap, and stale-output evals.
• Includes a smoke-test skeleton that can be adapted before production release.

Boundary: Not for replacing human review on destructive, financial, legal, medical, or safety-critical agent actions.