Grill With Docs

<what-to-do>

Interview me relentlessly about every aspect of this plan until we reach a shared understanding. Walk down each branch of the design tree, resolving dependencies between decisions one-by-one. For each question, provide your recommended answer.

Ask the questions one at a time, waiting for feedback on each question before continuing. Asking multiple questions at once is bewildering.

If a fact can be found by exploring the environment (filesystem, tools, codebase), look it up rather than asking me. The decisions, though, are mine — put each one to me and wait for my answer.

Do not act on the plan until I confirm we have reached a shared understanding.

A branch is closed when it is answered — or, rarely, when it is parked as an unresolved question (see below). Parking is the only sanctioned way to leave a branch open; never leave one silently unresolved.

When a question has options, explain every option so I can choose without looking anything up — see Presenting choices below.

</what-to-do>

<supporting-info>

Domain awareness

During codebase exploration, also look for existing documentation:

File structure

Most repos have a single context:

/
├── CONTEXT.md
├── docs/
│   └── adr/
│       ├── 0001-event-sourced-orders.md
│       └── 0002-postgres-for-write-model.md
└── src/

If a CONTEXT-MAP.md exists at the root, the repo has multiple contexts. The map points to where each one lives:

/
├── CONTEXT-MAP.md
├── docs/
│   └── adr/                          ← system-wide decisions
├── src/
│   ├── ordering/
│   │   ├── CONTEXT.md
│   │   └── docs/adr/                 ← context-specific decisions
│   └── billing/
│       ├── CONTEXT.md
│       └── docs/adr/

Create files lazily — only when you have something to write. If no CONTEXT.md exists, create one when the first term is resolved. If no docs/adr/ exists, create it when the first ADR is needed.

During the session

Presenting choices

When a question has options to pick from, do this for every option:

1. Show it in code first — a short, simplified snippet of what that option looks like in practice. Use the project's real names (files, functions, tables, domain terms) when you know them; otherwise a tiny made-up example. A few lines: just enough to see the difference, not a full implementation.
2. Explain it in plain words — one or two short sentences on what it means for the user: what gets easier, what gets harder, when you'd pick it.

Keep the language simple:

• No vague or buzzword-only terms ("more scalable", "cleaner", "decoupled", "idiomatic"). Say what actually happens instead.
• If a technical term is unavoidable, explain it in the same sentence.
• Make the options differ visibly — side by side, the snippets should make the trade-off obvious.

Then give your recommendation and the one-line reason. See EXAMPLES.md for what this looks like.

Challenge against the glossary

When the user uses a term that conflicts with the existing language in CONTEXT.md, call it out immediately. "Your glossary defines 'cancellation' as X, but you seem to mean Y — which is it?"

Sharpen fuzzy language

When the user uses vague or overloaded terms, propose a precise canonical term. "You're saying 'account' — do you mean the Customer or the User? Those are different things."

Discuss concrete scenarios

When domain relationships are being discussed, stress-test them with specific scenarios. Invent scenarios that probe edge cases and force the user to be precise about the boundaries between concepts.

Cross-reference with code

When the user states how something works, check whether the code agrees. If you find a contradiction, surface it: "Your code cancels entire Orders, but you just said partial cancellation is possible — which is right?"

Update CONTEXT.md inline

When a term is resolved, update CONTEXT.md right there. Don't batch these up — capture them as they happen. Use the format in CONTEXT-FORMAT.md.

CONTEXT.md should be totally devoid of implementation details. Do not treat CONTEXT.md as a spec, a scratch pad, or a repository for implementation decisions. It is a glossary and nothing else.

Park what the user genuinely cannot answer yet

Occasionally the grill hits a question the user simply cannot close right now — the decision belongs to someone else, depends on data that doesn't exist yet, or is waiting on something outside the codebase. The feature still has to get built. Rather than stalling the grill or letting the agent quietly guess, park the question and build a black box: a named placeholder, generic or empty, that can be swapped when the answer arrives.

This is rare. Expect zero parked questions in most sessions. It is not a valve for "the user hesitated" or "this is getting long".