Percerno Skillsbeta
← everything it can do · productivity

Grilling TEXT

Conducts a structured, iterative interview to stress-test a user's plan or idea by mapping a design tree of decisions. It asks targeted questions in rounds, provides recommended answers, and resolves dependencies until a shared understanding is reached.

7 uses this weeksource mattpococklicence MIT
Quality gate: FAIL · better answers · measured 2026-10-23
Blind-judged against the bare model on 2 requests: average lift -9.5 points (specificity, structure, method, fit).
won 0/2, lift -9.5, tokens ×3.15
Security scan: clean · 2026-10-24 · 2 files · ran in sandbox
no network calls, shell tricks, credential access or instruction tampering found in its files.

Try asking

Say it in your own words; the assistant finds this recipe by itself. Examples:

“Interrogér mig om min plan til at lancere det nye produkt i næste kvartal.”
“Jeg overvejer at skifte til en cloud-baseret løsning. Find svaghederne i min tankegang.”
“Hjælp mig med at teste mine antagelser om, at vi kan spare 20% på omkostninger ved at automatisere.”
“Grill me on my plan to migrate our database to PostgreSQL next month.”
“I want to launch a SaaS product in Q3. Stress-test my assumptions about the target market.”
“Help me think through the trade-offs of hiring a CTO versus outsourcing development. Ask me hard questions.”
Try this skill from Claude, ChatGPT or a chat on your own page. Free plan, no credit card.Get started free

Good for

stress-testdecision-makinginterviewplanningcritical-thinkingdesign-treeclarificationstrategy

Not for

Do not use this for simple factual queries or direct coding tasks where no strategic decision-making is required. It is not suitable for creative brainstorming where divergent thinking is preferred over convergent validation. Avoid using it for tasks that require immediate execution without prior planning or analysis.

Details

When it is the right choice

Use this skill when a user explicitly asks to be 'grilled' or wants to stress-test a specific plan, decision, or idea. It is ideal for complex scenarios where multiple dependent decisions need to be clarified before execution. It is also appropriate when a user has a vague concept and needs to flesh out the logical branches of their thinking.

The full recipe the assistant reads

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. 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.

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>

Word each question so "yes" accepts 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.

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; don't ask the user for anything you could look up yourself. Don't 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.

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.

Files in this skill (3)
  • LICENSE
  • SKILL.md
  • agents/openai.yaml
Where it comes from

Vendored from mattpocock/skills, commit 5beb35b040e3, path skills/productivity/grilling, licence MIT. The upstream licence file ships inside the skill.