socratic-grill

Clarify vague workplace plans by exposing hidden assumptions one question at a time.

1|Updated Mar 31, 2026
One-click install
npx skills add https://github.com/jiangding01/claude-tools --skill socratic-grill
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: socratic-grill
Source: https://github.com/jiangding01/claude-tools/tree/main/skills/socratic-grill
Command: npx skills add https://github.com/jiangding01/claude-tools --skill socratic-grill

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill helps you turn vague workplace ideas, plans, and disagreements into clear, decision-ready thinking. Instead of jumping straight to advice, it exposes hidden assumptions, identifies the real question, and keeps the conversation focused on the highest-leverage uncertainty.

Core Features & Use Cases

  • One-question-at-a-time questioning: Keeps the discussion disciplined so each reply sharpens the most important unknown.
  • Scenario routing: Adapts to product reviews, technical design, launch risk checks, data attribution, retrospectives, and cross-team conflict.
  • Assumption and evidence pressure-testing: Separates facts from guesses, challenges vague language, and asks for the evidence behind confident claims.
  • Cross-functional role lenses: Re-checks the same plan from PM, Dev, QA, Ops, SRE, Data, PMO, and Manager viewpoints to surface blind spots.
  • Convergence support: When the discussion is ready, it can help you collect the agreed points, risks, decisions, and next actions into a usable consensus format.

Quick Start

Tell the skill your specific plan, decision, or disagreement, and ask it to pressure-test the key assumptions one question at a time.

Frequently Asked Questions about socratic-grill

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I pressure-test hidden assumptions in a vague product plan?

Pressure-testing a vague plan requires exposing hidden assumptions through one-question-at-a-time Socratic questioning. This method separates facts from guesses, challenges vague language, and keeps discussion focused on the highest-leverage uncertainty to reach a decision-ready state.

How does Socratic questioning work for cross-functional disagreements?

Socratic questioning for cross-functional disagreements works by re-checking the same plan through different role lenses like PM, Dev, QA, Ops, and SRE. This surfaces blind spots and routes ambiguous inputs into a structured consensus canvas of agreed points and next actions.

What is the best way to review a PRD and identify missing evidence?

The best way to review a PRD for missing evidence is applying assumption pressure-testing to separate confirmed facts from unverified guesses. This approach challenges confident claims by asking for supporting evidence and clarifying vague language one question at a time.

Can I use assumption testing for technical design reviews and launch checks?

Yes, you can use assumption testing for technical design reviews and launch checks. Scenario routing adapts the questioning process specifically to launch risk checks and technical designs, exposing hidden risks and vagueness before converging on actionable decisions.

How do I facilitate a retrospective that turns vague conflicts into decisions?

To facilitate a retrospective turning vague conflicts into decisions, apply structured convergence support. This collects agreed points, risks, and next actions into a usable consensus canvas after pressure-testing the discussion through cross-functional role lenses.

When should I not use one-question-at-a-time questioning for workplace communication?

You should not use one-question-at-a-time questioning when a decision requires immediate emergency response or when the plan contains no ambiguous inputs or hidden assumptions to pressure-test, as the structured convergence process adds unnecessary overhead.