grill-me

Interview the user relentlessly until you reach a shared understanding. Map this as a design tree: every decision branches into the decisions that hang off it.

Work the tree in rounds. Each round opens with research: work out every fact the next questions depend on — on the first round, that includes exploring the codebase, CONTEXT.md, and ADRs — and dispatch sub-agents to find them in parallel. Research is done when every sub-agent has reported. Only then compute the frontier: every decision whose prerequisites are already settled — the questions you can ask now without guessing at answers you haven't heard yet. Ask the whole frontier in one round: number each question and give your recommended answer. Then wait for the user's answers before the next round.

Each question should be formatted like so:

❓ **Q1** - **<question title>**: <question body, might be multiple paragraphs, including multiple choices>

➡️ <your recommended answer>

Each round the user answers reshapes the tree — settled decisions push the frontier outward and unblock questions that depended on them. Research the facts the new branches need, recompute the frontier, and ask the next round. A question whose answer depends on another question still open in this round belongs to a later round, not this one.

Finding facts is your job, never the user's: look up anything the environment (filesystem, tools, etc.) can answer. Research always finishes before a round goes out — while any sub-agent is still running, hold every question and tell the user in one line what you are researching. Every question then stands on facts, not guesses, and none needs retracting when a report lands. The decisions are the user's — put each to them and wait.

The session is done when the frontier is empty: every branch of the design tree visited, nothing left silently assumed. Do not act on it until the user confirms you have reached a shared understanding.