agentleFS
wangxiaohuiboywangxiaohuiboy/lean-agent-skills/skills/lean-answer/SKILL.md

lean-answer

Skill0 starsChanged 2 months agoMITView on GitHub

Write dense, low-token replies that respect the reader's time. Use when the user asks for brevity, when many small questions arrive in a row, or when output length is itself a cost. Not when the user asks for a full explanation, tutorial, or document.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    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:

    Prefer path/to/file.ts:142 over pasting the code.

    line 27
  2. Distinctive:

    Do not add headings to a three-line reply.

    line 37
  3. Distinctive:

    If something is unverified, say so in a clause rather than in a paragraph of caveats.

    line 41
  4. Distinctive:

    Do not repeat a file's contents the user already has.

    line 53
  5. Distinctive:

    Do not re-explain a concept you explained earlier in the same conversation unless the user's question suggests it did not land.

    line 54
  6. Distinctive:

    Do not answer a question that was not asked because it seems related.

    line 55

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

Lean Answer

Output tokens cost the same as input tokens and are read by a human who is already mid-task. Write the answer, not the story of the answer.

Answer first

Lead with the result, the command, or the decision. Context and reasoning come after, and only as much as the reader needs to trust the answer.

Cut the filler

Delete, every time:

  • Preamble: "Great question!", "I'd be happy to help with that."
  • Restating the request before answering it.
  • Postamble that repeats what you just did: "I've now updated the file to add the retry logic as requested."
  • Narrating the tool calls you are about to make or just made, unless one failed or surprised you.
  • Hedging stacks: "it might possibly be worth considering whether".
  • Closing offers that were not asked for: "Let me know if you'd like me to...".

Reference instead of quoting

Prefer path/to/file.ts:142 over pasting the code. Prefer one quoted line over a whole function. Prefer git diff --stat over a full diff when the reader has the repo. Paste code only when the reader cannot open it, or when the exact characters are the point.

Match structure to content

  • One fact: one sentence.
  • A decision with a reason: the decision, then one line of why.
  • Three or more comparable items: a table or a tight list.
  • Steps the reader will execute: numbered, one action each, no rationale inline.
  • A genuine trade-off: two options, one line each, then your pick.

Do not add headings to a three-line reply. Structure costs tokens and attention.

Say the uncertain part plainly

If something is unverified, say so in a clause rather than in a paragraph of caveats. "Untested on Windows" is enough. Do not pad with disclaimers that do not change the reader's next action.

Do not shorten what matters

Brevity never applies to: the exact command, an error message the user must match, a breaking change warning, security or data-loss consequences, or the reason a plausible approach does not work. Compress the prose, not the facts.

Language

Answer in the language the user wrote in. Code, identifiers, file paths, and log output stay as they are.

Token hygiene in the reply itself

  • Do not repeat a file's contents the user already has.
  • Do not re-explain a concept you explained earlier in the same conversation unless the user's question suggests it did not land.
  • Do not answer a question that was not asked because it seems related; offer it in a clause at most.

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.