What problem does it solve? Creating a trustworthy domain adapter for a specialized field requires research, structured generation, and verification that weaker models often skip, producing unverified or overclaimed skill bundles. ## Core Features & Use Cases - Guided Domain Discussion: Interviews the user about use cases, trusted authorities, and red-lines before any generation begins, refusing licensure-gated or high-harm domains. - Four-Part Bundle Generation: Produces a step-by-step workflow with a mermaid flowchart, a domain adapter conforming to TEMPLATE.md, a trap fixture with GROUND-TRUTH.md, and a smoke eval comparing bare vs adapter runs. - Verification and Judging: Runs repo checks, smoke evals, and a fable-judge pass over the bundle's own claims before reporting results with declared debt. - Use Case: A team needs an adapter for a compliance research sector; the skill researches current regulations with fetched sources, writes the adapter and flowchart, builds a trap fixture for the sector's central fraud, and smoke-tests it. ## Quick Start Ask the AI to run /fable-domain for your sector, for example: create a fable domain adapter for regulatory compliance research.
