Grill with docs
Like /grilling, plus docs — a relentless interview to sharpen a plan or design that also creates ADRs and a glossary (CONTEXT.md) as decisions land.From its SKILL.md
npx -y skills add leobbaroni/cockpit --skill grill-with-docsAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.5 KB, as published. Nobody here has run it
Load BOTH the grilling skill and the domain-modeling skill (on Claude Code, invoke them via the Skill tool — actually invoke them, don't just apply their ideas; on other harnesses, open both skill files and follow them together), then run the interview on the plan given in the arguments.
grilling owns the interview: look before you ask, one question per turn with a recommended answer and a visible open-fork count, every fact tagged as stated / verified / assumed, and both gates run before the plan is handed over. Its TRACKS.md sets which slots this particular kind of work must decide; its PLAN-FORMAT.md sets where each section lands.
As decisions crystallise during the interview — not at the end — capture them per domain-modeling's rules: glossary terms into CONTEXT.md, hard-to-reverse choices into docs/adr/. A term pinned in CONTEXT.md is a (verified: CONTEXT.md) source for the rest of the interview; an ADR records what would have changed the answer.
When the interview concludes, persist the outcome per PLAN-FORMAT.md — SPEC.md for the contract, PLAN.md for the build order, Assumptions Ledger, and risks — and offer a done-condition the user can drive the build with (e.g. "complete PLAN.md build order; done when npm run build passes").