agentleFS
mehrad-dmmehrad-dm/mastermind/skills/clarify/SKILL.md

clarify

Skill23 starsChanged 2 months agoMITView on GitHub

Use the moment the user signals your last answer did not land: \"I did not follow that\", \"I don't understand\", \"say it simply\", \"simpler please\", \"more simply\", \"you lost me\", \"what does that mean\", \"too long\", \"explain again\", \"show me\", a confused question repeating something you already answered. Re-pitch what you just said, in plain words or as a picture. Not for a new question, and not for documenting a package: that is explain.

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

2 rules. 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 77 edited versions we read keep these.

  • 1

    Distinctive

    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. Very specific:

    Read ~/.mastermind/engineering/core/clarity.md for the register.

    line 12
  2. Distinctive:

    Do not restate the same structure in fewer words: If the first attempt was prose, the second one probably should not be.

    line 66

“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

MasterMind: Clarify

Your last answer failed. That is the fact to accept before writing anything else. Do not defend it, do not repeat it louder, and do not simply cut its length: a shorter version of an unclear thing is still unclear.

Read ~/.mastermind/engineering/core/clarity.md for the register. Use the project's own terms from .mastermind/brief.md if it exists.

Work out what actually failed

One of four things went wrong. The repair differs for each, so name it before you rewrite.

What failedThe repair
Missing context: you started in the middleLead with why this exists, then what it does
Vocabulary: a term they do not holdName it once in their words, then use one word for it
Shape: a structure prose cannot carryDraw it. See below
Volume: correct, and too much at onceOne claim, then stop and ask what they want next

Two forms. Pick, do not offer a menu

Words. Lead with the point. Short sentences, active voice, one instruction each, no hedging. State what you would do and why, in the order they need it, not the order you discovered it.

A picture, when the thing has a shape: how parts connect, what calls what, what changed, what a sequence does, what the data looks like. Reach for the lightest form that carries it:

  • a file layout: a shallow tree, one line of responsibility per entry, for where things live
  • a call stack: an indented call tree, for what happens in what order
  • a component tree: the same for a UI, keeping the state and the boundaries, dropping the rest
  • types and signatures: the shape of the data, which is often the whole answer
  • pseudocode: for an algorithm, where real code carries noise the reader does not need
  • a table: for a comparison or a set of options
  • a diff, and not only of code: mark a call tree, a file layout or a state machine with + and - to show a change against a shape the reader already holds
  • a mermaid diagram: state and sequence diagrams, only when the relationships genuinely need one. Nine boxes at most; past that, draw an overview and a detail. Remove every box, arrow and label whose removal loses nothing, and give emphasis to one or two elements, never four

Plain text beats a rendered diagram: it works in every terminal and every tool, and it is faster to read. Reach for HTML only when the user asks for it.

Then stop. End with one short question that finds the next gap: "is it the X part or the Y part?" A second wall of text is the same failure again.

Close the loop

Each time this fires you have found a place MasterMind writes badly, and that is worth more than the one repair. When the trigger was your own writing rather than a genuinely hard subject, add a line to the brain's own .mastermind/journal.md, the file mastermind wrong-log reads, in its format so the log shows it: <date> · wrong · <what you said that did not land> · caught by the user not following · <the shorter form that worked>.

Gotchas

  • Do not fire on a new question. "What does that mean" about your last answer is this skill. "What does CORS mean" is a question: answer it.
  • Do not lose a qualifier while simplifying. Dropping "only on the first install" makes a sentence shorter and wrong. Cut decoration, never conditions.
  • Do not restate the same structure in fewer words. If the first attempt was prose, the second one probably should not be.
  • Never say "as I said above". They read it. It did not 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.