git:20261015.0b3e11b to git:20261020.b184c85

3 added, 2 removed. Audit A to A.

---

name: grilling

description: Grill the user relentlessly about a plan, decision, or idea until every branch of it is settled. Use when the user wants to stress-test their thinking, says "grill me", "poke holes in this", "interrogate this plan", or asks to be questioned before building something.

---

# Grilling

Interview the user relentlessly until you reach a shared understanding. Credit and license:

`LICENSE.md` in this folder.

Map the plan as a **design tree**: every decision branches into the decisions that hang off it.

## Rounds

Work the tree in **rounds**. The **frontier** is 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.

+ the whole frontier in one round: number each question and give your recommended answer. Word

+ each question so "yes" accepts your recommended answer, so agreeing never means answering "no".

+ Then wait for the user's answers before the next round.

Format a round like so:

```

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

➡️ <your recommended answer>

---

❓ **Q2** - **<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. 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.

## Facts are yours, decisions are theirs

Finding _facts_ is your job, never the user's. When a frontier question needs a fact from the

environment (the code, the docs, a config, a tool's output), look it up: hand it to a background

subagent if you have one, otherwise read it yourself. Never ask the user for anything you could

look up. Don't block on a lookup: a running search is an unsettled prerequisite, so only the

questions downstream of it wait; ask the rest of the frontier now.

The _decisions_ are the user's: put each one to them and wait.

## Done

The session is done when the frontier is empty: every branch of the design tree visited,

nothing left silently assumed. Close with the settled decisions as a short list. Do not act on

them until the user confirms you have reached a shared understanding.