agentleFS
ychampionychampion/ystack/skills/grill/SKILL.md

grill

Skill4 starsChanged 2 days agoMITView on GitHub

Stress-test a plan or idea with pointed questions asked one at a time, each with a recommended answer, until more answers would not change the work; then write a short spec and a ticket list with what blocks what. The short version: use for "poke holes in this plan and give me tickets", "quick stress-test before I build", or before a build whose request leaves real choices open. For a long, open-ended interview with no spec at the end ("grill me", "grill me relentlessly"), use `grilling`.

89 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use write-skill, root-cause and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • NoaiRoxSkill

    write-skill

    Use when creating a new skill, editing an existing skill, or verifying a skill works before deployment.

    Also in 5 projects that use this

  • Ingramradical235Skill

    root-cause

    Structured root cause analysis using WHERE (MECE decomposition) then WHY (causal structure diagrams) pipeline. Includes coin-flip detection, logical-leap detection, and vicious cycle intervention. Use when buffer enters YELLOW/RED, when metrics stall, or when Actions execute but don't produce results.

    Also in 4 projects that use this

  • Neko-Catpital-LabsSkill

    prove-it

    Shared evidence rule for investigating, diagnosing, debugging, proving, or explaining what happened or why for any problem or situation: do not assert something is fixed, working, passing, merged, or running unless you personally observed it this turn and can show what you saw; do not assert it was caused by X until a repro and controlled isolation prove that cause.

    Also in 4 projects that use this

  • sahinSkill

    execute

    Also in 5 projects that use this

  • assisjpSkill

    file-pr

    Prepare and open or update an authorized pull request with a concise problem-first description and verifiable evidence. Use when the user asks to create, file, open, or update a PR. Do not use merely to review a diff or when publication is not authorized.

    Also in 3 projects that use this

  • Arman-KudaibergenovSkill

    write-plan

    Этот скилл MUST быть вызван когда есть design.md и нужно разбить доработку на атомарные задачи (2-5 мин) с точными путями файлов и критериями проверки в tasks.md. SHOULD также вызывать как промежуточный шаг между brainstorm и subagent-dev. Do NOT использовать для обсуждения идеи — используй brainstorm; для выполнения задач — используй subagent-dev.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 70 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the plan, the issue and the code it touches.

    line 28
  2. Distinctive:

    Write, privately, every decision the plan depends on.

    line 34
  3. Distinctive:

    Ask the highest-ranked open decision first.

    line 39
  4. Distinctive:

    Wait for the answer. Then re-rank: an answer often settles or creates other decisions.

    line 48
  5. Distinctive:

    Stop when every decision left would not change what gets built, or the user says stop.

    line 56
  6. Distinctive:

    Every question carries a recommended answer, so "yes" is a valid reply.

    line 85
  7. Distinctive:

    Ask about decisions, never about facts you could check.

    line 86
  8. Distinctive:

    No spec until the questions stop changing the work.

    line 87

“Common” is measured: kept by at least 6 in 10 of the 70 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Grill

A plan fails on the question nobody asked. This skill asks those questions, one at a time, and turns the answers into something an agent can build from.

When it applies

  • A build or change where two sensible readings would lead to different work.
  • A plan, design or idea the user wants challenged before committing to it.
  • A vague issue or request that needs to become a spec and tickets.

Skip it when the request is already precise. Build instead.

Procedure

1. Read before you ask

Read the plan, the issue and the code it touches. Any fact you can look up (what the code does now, which library is in use, what a config holds) is yours to find, not the user's to answer. Use a subagent for the lookups if your host has them.

2. List the open decisions

Write, privately, every decision the plan depends on. Rank them by how much work changes with the answer. Drop any whose answers would all lead to the same work.

3. Ask one question at a time

Ask the highest-ranked open decision first. Each question:

Q3. <the decision, one sentence>
Why it matters: <what changes in the work, one line>
Options: A) ... B) ... C) ...
Recommended: B, because <one line>

Wait for the answer. Then re-rank: an answer often settles or creates other decisions. Never ask something a previous answer already decided.

Push back once when an answer conflicts with an earlier one or with the code. Name the conflict and ask which wins.

4. Stop at the right point

Stop when every decision left would not change what gets built, or the user says stop. Usually that is 4-10 questions. Say "That's enough to build from" and move on.

5. Write the spec

Goal: <one sentence: who gets what>
Decisions: <each answer, one line each>
Out of scope: <what this will not do>
Done when: <checkable acceptance criteria, one per line>
Open: <anything still unknown, and the default you'll use>

6. Write the tickets

Split the work into tickets an agent can finish in one session. Prefer thin slices that work end to end over one ticket per layer.

T1  <title>  blocked by: none   done when: <check>
T2  <title>  blocked by: T1     done when: <check>
T3  <title>  blocked by: none   done when: <check>

Say which tickets can start now in parallel. Write them to a file or the user's tracker only if asked.

Rules

  • One question per message. A wall of questions gets skimmed.
  • Every question carries a recommended answer, so "yes" is a valid reply.
  • Ask about decisions, never about facts you could check.
  • No spec until the questions stop changing the work.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.