--- name: grilling description: Model-invoked, Pi-native reference for relentlessly stress-testing a plan, decision, or idea through focused questions. --- # /grilling - shared-understanding decision interviews Every message to the operator follows the `operator-communication` skill: load it before your first message. If a `tw_*` tool or the workflow misbehaves, refuses you, or otherwise surprises you, invoke the `workflow-failure` skill before improvising. Interview the user relentlessly until you reach a shared understanding. Map this as a **design tree**: every decision branches into the decisions that depend on it. The goal is shared understanding, not a performance of questioning. Preserve the user's decisions, expose trade-offs, and do not silently invent answers. ## Design tree and frontier 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 have not 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. A question that depends on another question still open in this round belongs to a later round. For every frontier question, be focused and provide a concrete recommended answer. Do not answer on the user's behalf. Format a round like this: ```text ❓ **Q1** - ****: ➡️ Recommended answer: --- ❓ **Q2** - ****: ➡️ 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 after each answer. ## Facts, ordering, and recording Finding _facts_ is your job, never the user's. When a frontier question needs a fact from the environment (filesystem, tools, etc.), dispatch a sub-agent to find it; do not ask the user for anything you could look up yourself. Do not block on it: a running exploration is an unsettled prerequisite, so only the questions downstream of it wait for the sub-agent to report; ask the rest of the frontier now. The _decisions_ are the user's: put each to them and wait. Order questions by prerequisites: settle parent decisions before dependent choices. After each round, record the user's answer in the relevant task or planning artifact in the user's terms. Keep a decision index when the current Wayfinder or task resource provides one. Preserve settled decisions; do not re-ask them. For each settled decision, retain important rejected options, trade-offs, constraints, rationale, and downstream consequences. If an answer creates work precise enough to state, hand it off to wayfinder (for planning decisions) or to-spec (for implementation decisions), not by duplicating task routing here. ## Completion gate Continue until the frontier is empty: every branch of the design tree has been visited, every prerequisite is settled, and nothing remains silently assumed. Then summarize the decision, alternatives, trade-offs, constraints, and consequences and ask the user to confirm that shared understanding has been reached. Do not act on the plan, mark a grilling task complete, or hand off implementation until that explicit confirmation. Keep task-specific resources as the operational adapters for their own artifacts and handoffs.