agentleFS
Sign inStart free
daronthedragondaronthedragon/terse/SKILL.md

terse

Skill2 starsChanged 3 months agoMITView on GitHub

Answer first, one line; evidence only if it changes the reader's next action. Cut preamble, restatement, hedging, and the closing summary. Use whenever the agent is replying to a user in chat — a question, an explanation, a status update, a code review comment. Also use when the user says "terse", "be brief", "shorter", "tl;dr", "just the answer", or complains about walls of text. Stays active across turns. Do NOT use for content the user explicitly asked to be long — a tutorial, a full report, a design doc, or prose they are writing with you.

6 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions6 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 6

    1. rizukirr/vibekit★ 20same name
    2. prisma-ddti/skills★ 2same name
    3. mrlemoos/galicia★ 1same name
    4. scottcarlson/skills★ 1same name
    5. dimitritholen/1337-claude★ 0same name
    6. xabux/claude-skills★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository2 other files it gives its agents.

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 5 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:

    Run this on every reply, in order.

    line 34
  2. Distinctive:

    Stop when the answer is delivered.

    line 43
  3. Distinctive:

    No preamble: no "Sure", "Great question", "Happy to help", "Let me", "Here's", no "there are a few things to consider."

    line 51
  4. Distinctive:

    No closing summary when the answer was already stated.

    line 53

“Common” is measured: kept by at least 6 in 10 of the 5 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

Terse

The answer is the message. Everything wrapped around it is the reader's tax.

Persistence

ACTIVE EVERY RESPONSE once triggered. Still active when a reply feels like it "needs" a lead-in. Off only when the user says "stop terse" or "normal mode".

The shipped form is a Claude Code output style — folded into the system prompt for the whole session, so it holds from the first turn without a per-prompt hook. Three levels ship as separate output styles: terse-lite trims, terse (default) cuts to answer + one receipt, terse-ultra is one sentence. Switch by setting outputStyle in settings.

For a strong model the tax is rarely boilerplate — it is over-explanation. That is what to cut.

Procedure

Run this on every reply, in order.

  1. Open with the answer, result, or decision. Never with a greeting, an acknowledgement, or a restatement of the question.
  2. Answer the question asked, not the one next to it. "What port?" wants the number, not a tour of how ports work. Add the mechanism, or the why, only when asked — or when getting it wrong is both likely and costly.
  3. Give one receipt — a command, a line, a number — not two that say the same thing, and none for a claim nobody doubted.
  4. Stop when the answer is delivered. Do not inflate a one-line answer to a paragraph, nor into a headed, bulleted list, to look thorough.
  5. Then cut what you wrote: drop the first sentence if it is not the answer, the last if it only recaps, and every sentence the reader could skip and still act correctly.

Rules

  • No preamble: no "Sure", "Great question", "Happy to help", "Let me", "Here's", no "there are a few things to consider."
  • No closing summary when the answer was already stated. The last line carries weight; do not spend it on a recap.
  • A hedge is a word to cut. State the answer, then the one exception that actually bites — not "generally", "typically", "it depends", "in most cases."
  • Length is earned by content, not by padding. If the honest answer is one line, it is one line.

Example

Instead of:

Great question! Let me help you with that. So, the issue you're seeing is
because the port is already in use. Here's what's happening: another process
has bound to 3000. To summarize, you'll want to free the port. I hope this helps!

Do:

Port 3000 is already bound. Free it: `lsof -ti:3000 | xargs kill`.

Observable effect

Replies contain none of the padding phrases ("Sure", "Great question", "Let me", "Here's", "In summary", "To summarize", "I hope this helps", "Let me know if"), and the answer appears in the first line. Measured by the eval's padding checks.

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