agentleFS
charliehzmcharliehzm/medharness/.claude/skills/ask-questions-if-underspecified/SKILL.md

ask-questions-if-underspecified

Skill85 starsChanged 6 months agoApache-2.0View on GitHub

Use this skill when scope ambiguity blocks progress in any SOP step (PRD, TDD, design, task decomposition). Produces a tight set of clarifying questions (max 5) that, once answered, unblock the next step. Avoids the failure mode where Agents proceed on assumption and produce wrong work. Chinese trigger examples: "需求不清", "需要澄清", "提问澄清", "范围不明", "信息不足要澄清", "ask the user". Do NOT use for exploration (use research), do NOT use to delay obvious work. Success = ≤ 5 questions, each decision-shaping, each tied to a specific PRD section.

15 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use llm-application-dev, audit-library-health 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.
  • pepebot-spaceSkill

    llm-application-dev

    Building applications with Large Language Models - prompt engineering, RAG patterns, and LLM integration. Use for AI-powered features, chatbots, or LLM-based automation.

    Also in 4 projects that use this

  • MoizIbnYousafSkill

    audit-library-health

    Use when checking the overall health of a skills library. Run doctor, validate, check for stale skills, and verify generated docs are in sync.

    Also in 3 projects that use this

  • MoizIbnYousafSkill

    browse-and-evaluate

    Use when exploring the ai-agent-skills catalog to find, compare, and evaluate skills before installing. Always use --fields to limit output size and --dry-run before committing to an install.

    Also in 3 projects that use this

  • pepebot-spaceSkill

    code-documentation

    Writing effective code documentation - API docs, README files, inline comments, and technical guides. Use for documenting codebases, APIs, or writing developer guides.

    Also in 4 projects that use this

  • +8 more from MoizIbnYousaf

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 15 edited versions we read keep these.

  • 3

    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:

    Only ask decision-shaping questions.

    line 29
  2. Distinctive:

    Keep at most 5 of type decision-shaping or compliance-touching.

    line 46
  3. Distinctive:

    Never: ask "is this OK" / "should I proceed" — those are not clarifications.

    line 49

“Common” is measured: kept by at least 6 in 10 of the 15 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

Ask Questions If Underspecified

The "stop and clarify" gate. Many AI Coding failures trace back to "Agent assumed X, X was wrong, 4 hours wasted." This skill catches that early.

Core principle

Only ask decision-shaping questions. If the answer changes nothing about what gets built, don't ask.

Question taxonomy

For each candidate question, classify:

TypeAsk?Example
Decision-shapingYES"Should L4 PHI be stored encrypted at rest or only redacted?"
PreferenceOnly if 2 acceptable paths"Frontend React or Vue?"
TriviaNO"What color is the button?" (decide and move on)
Compliance-touchingALWAYSAnything touching tier / model / audit

Workflow

  1. List unknowns surfaced from the upstream artifact (PRD / TDD / design).
  2. Classify each per the taxonomy.
  3. Keep at most 5 of type decision-shaping or compliance-touching.
  4. Phrase each question so the answer is bounded (multiple-choice or numeric).
  5. Format with AskUserQuestion tool when available; otherwise inline.
  6. Never ask "is this OK" / "should I proceed" — those are not clarifications.

Output

A list of ≤ 5 questions, each with:

  • Header (≤ 12 chars chip label)
  • Why it matters (one line)
  • 2-4 mutually exclusive options (or "numeric input")
  • Recommended default

Common failure modes

  1. Question dump — asking 15 questions exhausts the user. Cap at 5; everything else is your call.
  2. Open-ended phrasing — "What do you think about X?" produces non-actionable answers. Always bound.
  3. Asking after committing — questions came too late, work already started. Run this skill earlier in the SOP.

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.