agentleFS
Sign inStart free
PeanutGTPeanutGT/NutsSkills/skills/clarify/SKILL.md

clarify

Skill3 starsChanged 19 days agoApache-2.0View on GitHub

Use when a request is vague, open-ended, or a new feature idea (I want to build X, add support for Y), before writing a spec, plan, or code. Decides whether questions are needed at all, then interviews one question at a time until goals, constraints, and edge cases are unambiguous.

80 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use quieter, arrange 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.
  • HK-hubSkill

    quieter

    Tones down visually aggressive or overstimulating designs, reducing intensity while preserving quality. Use when the user mentions too bold, too loud, overwhelming, aggressive, garish, or wants a calmer, more refined aesthetic.

    Also in 29 projects that use this

  • HK-hubSkill

    arrange

    Improve layout, spacing, and visual rhythm. Fixes monotonous grids, inconsistent spacing, and weak visual hierarchy. Use when the user mentions layout feeling off, spacing issues, visual hierarchy, crowded UI, alignment problems, or wanting better composition.

    Also in 14 projects that use this

  • zwy123zwySkill

    normalize

    将用户随意输入的提示词规范化为结构清晰、便于大模型执行的标准化提示词。适用于需要整理、统一提示词格式时调用。

    Also in 19 projects that use this

  • owshq-mecSkill

    layout

    Improve layout, spacing, and visual rhythm. Fixes monotonous grids, inconsistent spacing, and weak visual hierarchy. Use when the user mentions layout feeling off, spacing issues, visual hierarchy, crowded UI, alignment problems, or wanting better composition.

    Also in 8 projects that use this

  • +8 more from HK-hub

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 77 edited versions we read keep these.

  • 4

    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:

    Iron law: No spec, plan, or code while a question that would change what gets built is still open.

    line 14
  2. Distinctive:

    Ask yourself: what do I know about what they want, and why?

    line 19
  3. Distinctive:

    Look before you ask: Read the relevant code, docs, and recent commits.

    line 24
  4. Distinctive:

    Lay out the approaches: When there is more than one sensible way to build it, describe two or three with their trade-offs, recommend one, …

    line 32

“Common” is measured: kept by at least 6 in 10 of the 77 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

Clarify requirements

Iron law: No spec, plan, or code while a question that would change what gets built is still open.

Building the wrong thing well is still building the wrong thing. Asking about things that don't matter wastes the user's time just as surely.

0. Triage first

Ask yourself: what do I know about what they want, and why?

  • Everything that matters is in the request ("make the button blue"): do it and say what you did. No interview.
  • Something non-trivial is missing: continue below.

Process

  1. Look before you ask. Read the relevant code, docs, and recent commits. Never ask what the repo can tell you.
  2. Restate the goal in one or two sentences: who it is for, what problem it solves, how success is measured.
  3. One question per message. Each question:
    • is about one decision, specific to their project;
    • gives your recommended answer and why;
    • offers 2–4 concrete options when there are clear alternatives.
  4. Walk the decision tree. Settle parent decisions before children.
  5. Probe the edges: empty and huge inputs, failure modes, permissions, concurrency, existing data, undo.
  6. Lay out the approaches. When there is more than one sensible way to build it, describe two or three with their trade-offs, recommend one, and let the user pick.
  7. Play it back in chunks. As soon as one part is clear, describe it back briefly. Mark what the user said as theirs and what you filled in as your proposal, with your reason. A default they never heard is one they could not reject.
  8. Size the work in plain words (small change / project) and name what is out of scope.
  9. Summarize the decisions and get a clear "yes" before moving on.

Question checklist

  • Goal and success measure
  • Users / callers and their main scenarios
  • Inputs, outputs, and error behavior
  • Constraints: performance, security, compatibility, deadlines
  • Existing code or patterns this must fit
  • Explicitly out of scope

Rationalizations to reject

ExcuseAnswer
"The intent is obvious"Then confirming it takes one message.
"I'll pick sensible defaults"Good — say them out loud as proposals. Silent defaults are guesses.
"I'll ask everything at once to save time"A wall of questions gets skimmed. One sharp question gets answered.
"This is too small to ask about"Then triage says do it. Don't interview for a one-liner.

Done when

  • The user approved a written summary of decisions, including which approach
  • No open question would change what gets built

Next: spec (or straight to plan for a small change).

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