: <body, may list multiple choices> rec: <your recommendation> ``` Fuzzy or overloaded term surfaces mid-interview? Coin a canonical term and hand it to the calling loop to write into its glossary — do not write it yourself (you own no artifacts). ## BOUNDARIES - MUST NOT write any artifact (plan folder, PRD, glossary, file). - MUST NOT decide for the user — recommend, then wait. - MUST NOT block a round on a fact lookup — ask the rest of the frontier. - MUST NOT re-ask settled decisions. - MUST NOT act on the interview until the user confirms shared understanding.

# Grilling > Relentless interview primitive: map the design tree, work the frontier in rounds — every question whose prerequisites are settled gets asked in one round, numbered, each with a recommended answer. Facts are the agent's job, never the user's. Use when any loop needs to sharpen a fuzzy idea by interview (grill, triage, deepen, wayfinder), or when the user asks to be grilled, stress-tested, or challenged. - Skill: `arendon1/grilling` (Agent Skill) - Install (CLI): `npx skillmds@latest add arendon1/grilling` - Raw SKILL.md: https://api.skillmd.com/api/skills/arendon1/grilling/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: arendon1 (https://skillmd.com/u/arendon1) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/arendon1/grilling --- # grilling — the interview primitive Interview engine. Runs the conversation that turns fuzz into decisions. Owns nothing, writes nothing: no PRD, no CONTEXT.md, no plan folder. The calling loop owns artifacts. Pure protocol, reusable by any loop that must sharpen an idea by interview. ## THE MODEL Map the discussion as a **design tree**: every decision branches into the decisions that hang off it. - **Frontier** — every decision whose prerequisites are settled: the questions askable NOW without guessing answers you have not heard yet. - **Round** — ask the WHOLE frontier in ONE round. Number each question, give your recommended answer for each. Wait for all answers. - **Recompute** — answers reshape the tree: settled decisions push the frontier outward and unblock questions that depended on them. Next round asks the new frontier. - A question whose answer depends on another question still open this round belongs to a LATER round, not this one. Done when the frontier is empty: every branch visited, nothing silently assumed. Do not act until the user confirms shared understanding. ## FACTS ARE YOUR JOB, DECISIONS ARE THE USER'S - Frontier question needs a fact from the environment (filesystem, tools, external source)? Look it up yourself. NEVER ask the user anything you can find. - Expensive lookup? Delegate to a background helper and DO NOT block the round: a running exploration 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 to them and wait. ## QUESTION FORMAT Each question carries your recommended answer so the user can grunt "yes" and move: ``` Q — : <body, may list multiple choices> rec: <your recommendation> ``` Fuzzy or overloaded term surfaces mid-interview? Coin a canonical term and hand it to the calling loop to write into its glossary — do not write it yourself (you own no artifacts). ## BOUNDARIES - MUST NOT write any artifact (plan folder, PRD, glossary, file). - MUST NOT decide for the user — recommend, then wait. - MUST NOT block a round on a fact lookup — ask the rest of the frontier. - MUST NOT re-ask settled decisions. - MUST NOT act on the interview until the user confirms shared understanding.