agentleFS
Sign inStart free
willianbswillianbs/skills/requirements-grill/SKILL.md

requirements-grill

Skill0 starsChanged 4 months agoMITView on GitHub

Adversarially clarifies product and engineering requirements before Spec Validator. Use when asks are vague or high-stakes. Emits GRILL_NOTES. Never invents product decisions as facts or starts implementation.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Prefer questions that unlock decisions over essays.

    line 38
  2. Distinctive:

    Do not silently fill gaps.

    line 43
  3. Distinctive:

    Never pretend product chose an option they did not.

    line 88
  4. Very specific:

    Never start coding from GRILL_NOTES alone.

    line 89
  5. Distinctive:

    Never shame the asker.

    line 90

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

Purpose

Stress-test an ask with pointed questions and contradiction checks so SPEC_STATUS is earned, not assumed.

When to Use / When NOT to Use

Use when: vague tickets, conflicting stakeholders, HighRisk features, user says “grill me” / “challenge the requirements”.

Do not use when: already Ready SPEC_STATUS; pure bug with clear expected behavior; user wants immediate coding only (offer handoff).

Preconditions

An ask exists (even messy). If empty, ask for the one-sentence goal first.

Inputs / Outputs

Inputs: raw ask/ticket/PRD draft, optional CONTEXT_PACK.

Outputs: GRILL_NOTES (feeds spec-validator).

Upstream / Downstream

Upstream: user, engineering-mentor, engineering-os.

Downstream: spec-validator, delivery-planner, threat-modeler (if HighRisk).

Core Principles

  1. Adversarial but respectful — attack ambiguity, not people.
  2. Prefer questions that unlock decisions over essays.
  3. Surface hidden assumptions, non-goals, and edge cases.
  4. Force measurable acceptance criteria candidates.
  5. Separate ProductDecisionNeeded vs EngineeringDecisionNeeded.
  6. Cap to the smallest question set that unblocks Ready.
  7. Do not silently fill gaps.

Process

  1. Restate the ask in one sentence; confirm.
  2. Probe: users, success metric, non-goals, abuse cases, data, scale, deadline, platforms.
  3. Contradiction hunt — flag conflicts.
  4. Edge-case battery — empty, failure, permissions, concurrency, migration.
  5. Draft candidate AC IDs (provisional) for spec-validator.
  6. List blocking questions vs nice-to-have.
  7. Decision: ready for spec-validator or still Blocked on product answers.

Evidence Requirements

Quote vague phrases. Do not invent stakeholder answers.

Stop Conditions / Failure Modes

ConditionAction
Critical product answers missingBlock handoff to planner/implementer
Ask is already preciseShort grill → pass to spec-validator

Severity + Confidence

Missing answers on auth/money/PII → High blockers.

Output Contract

## GRILL_NOTES
Restatement: ...
Assumptions (unconfirmed): ...
Blocking questions: ...
Non-blocking questions: ...
Candidate ACs: AC-1...
Decision: Proceed | ProceedWithConditions | Revise | Block

Handoffs

spec-validator (required next), threat-modeler, engineering-mentor.

Never

  • Never pretend product chose an option they did not.
  • Never start coding from GRILL_NOTES alone.
  • Never shame the asker.

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.

Check your setup, free