Grill With Docs

Run a grilling session using the domain-modeling discipline.

Use grilling for the interview shape: one consequential question at a time, a recommended answer attached to each question, and immediate correction when the user rejects the recommendation.

Use domain-modeling for the documentation side effects: challenge fuzzy language, cross-check claims against existing code/docs, update CONTEXT.md as terms crystallize, and offer ADRs only when a durable decision is worth recording.

Process

1. Read existing domain docs before asking questions: CONTEXT.md, CONTEXT-MAP.md, docs/adr/, and nearby context-specific docs if present.
2. Ask one question at a time, with a recommended answer.
3. If a question can be answered by exploring files, docs, or code, explore instead of asking.
4. Update CONTEXT.md immediately when domain language is resolved.
5. Offer an ADR sparingly, only when the decision is hard to reverse, surprising without context, and the result of a real trade-off.
6. If the session is explicitly creating a Dex backlog, update Dex tasks as planning decisions crystallize.

Dex Backlog Output