agentleFS
ohongohong/agent-skills/writing-with-smart-brevity/SKILL.md

writing-with-smart-brevity

Skill4 starsChanged 2 days agoMITView on GitHub

Rewrites long-winded, padded or AI-sounding text into Smart Brevity style (Axios / VandeHei, Allen & Schwartz) - main point first, why it matters, ranked support, clear next step - without losing facts, caveats or warmth. Use when the user asks to tighten, shorten, edit, punch up or "make concise" an email, memo, update, announcement, Slack post, summary or doc, mentions Smart Brevity or Axios style, or complains that writing is too long, rambling, buried, or sounds like AI slop.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 1

    1. ohong/skills★ 4100% same

    Edited · 0

    None found.

    Wrote their own version · 0

    None found.

    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 repository10 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 8 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    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. Distinctive:

    Go deeper: only the support the reader needs, ranked by importance, then an optional path to more.

    line 17
  2. Distinctive:

    Do these in order. Structure first, words last.

    line 21
  3. Distinctive:

    One big thing: Write the single sentence you'd keep if everything else vanished.

    line 24
  4. Distinctive:

    Protect: Before cutting, mark what would change a reasonable reader's interpretation: numbers, names, dates, denominators, qualifiers …

    line 26
  5. Distinctive:

    Cut: Delete throat-clearing, things the reader knows, repetition across layers, empty phrases, and ritual closers.

    line 27
  6. Distinctive:

    Verify: Read it aloud and fix stumbles.

    line 30
  7. Distinctive:

    Keep connected reasoning as prose: If items link with because / but / so, those links are the content.

    line 44
  8. Distinctive:

    Leave good writing alone: If the draft is already ordered, qualified and tight, make the smallest edit, or none.

    line 47
  9. Very specific:

    Don't fabricate: If the draft lacks an owner, deadline or stake you need, flag it as [owner?] instead of inventing it.

    line 49

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Writing with Smart Brevity

Smart Brevity means the writer does the prioritizing so the reader doesn't have to. It's short, not shallow: length falls out of good judgment; it's never the goal. Compress reader effort, not truth.

Governing test: from the first screen alone, can the reader get the main point and why it matters, then choose how much deeper to go?

The Core 4

  1. Tease (headline / subject): specific news, not a label. Strongest word first, under 10 words ("Office closes Oct. 17", not "Office transition update").
  2. Lede: the one big thing in one sentence. It pays off the tease and doesn't repeat it.
  3. Why it matters: 1-2 direct sentences on the consequence for this reader. Skip it if the stakes are obvious.
  4. Go deeper: only the support the reader needs, ranked by importance, then an optional path to more.

Workflow

Do these in order. Structure first, words last; polishing a sentence you should delete is wasted work.

  1. Reader. Name them and the outcome: "After reading, [reader] should [know / decide / do X]." Cut material aimed at showing the writer's effort.
  2. One big thing. Write the single sentence you'd keep if everything else vanished. If two things compete, pick one or split the message.
  3. Order. Point → why it matters → strongest support → caveats → action / depth. Background and chronology go last or go away.
  4. Protect. Before cutting, mark what would change a reasonable reader's interpretation: numbers, names, dates, denominators, qualifiers ("may", "in this sample"), conditions, risks, counterevidence. These can move or compress, never vanish.
  5. Cut. Delete throat-clearing, things the reader knows, repetition across layers, empty phrases, and ritual closers. Use plain words, active verbs, and subject-verb-object order.
  6. Scan layer. One idea per paragraph. Bullets only for 3+ parallel items. Bold the one phrase per section that deserves a second look. Add an axiom label only when it names what a block does.
  7. Close. End with who does what by when. If nothing's needed, end when the information ends.
  8. Verify. Read it aloud and fix stumbles. Try deleting the last sentence. Diff against the source: same facts, same certainty, same emphasis? If not, repair it.

AI slop to remove

  • Openers: "Great question!", "I hope this finds you well", "I wanted to reach out", "In today's fast-paced world", restating the question.
  • Closers: "Let me know if you have any questions", "I hope this helps", "In summary..." recaps of what was just said.
  • Filler: "It's worth noting", "It's important to remember", "at this point in time", "in order to", "a wide range of", "plays a crucial role".
  • Inflated words: utilize, leverage, facilitate, delve, robust, seamless, landscape, journey, unlock, empower, pivotal, holistic.
  • Tics: "not just X, but Y", reflexive triplets, stacked hedges ("could potentially perhaps"), "-ing" tails that add fake significance ("..., highlighting the importance of").
  • Format cargo cult: a header on a 3-sentence answer, bullets of clipped fragments, bold every third phrase, decorative emoji, every Axios label used whether or not it fits.

Don't overcorrect

  • Keep caveats. A punchier claim that's now wrong isn't brief, it's wrong. Never raise certainty or scope.
  • Keep connected reasoning as prose. If items link with because / but / so, those links are the content; bullets break them.
  • Keep warmth in hard news (layoffs, grief, mistakes). Lead with the point, but say it like a person would face to face.
  • Keep precise terms (legal, medical, technical) the reader needs. Translate jargon they don't.
  • Leave good writing alone. If the draft is already ordered, qualified and tight, make the smallest edit, or none.
  • No made-up rules. There's no word-count ceiling per sentence or paragraph, no ban on passive voice, and no required "Why it matters" label.
  • Don't fabricate. If the draft lacks an owner, deadline or stake you need, flag it as [owner?] instead of inventing it.

Formats

Carry the hierarchy, not the Axios template. Use only the fields that earn a place: "Client approved. I'll send the contract by 3pm." can already be optimal.

Email / Slack     Subject: [the news or ask]
                  [Main point, one sentence.]
                  Why it matters: [consequence for reader.]
                  [Only what's needed to understand or act.]
                  Next: [who] will [do what] by [when].

Exec update       1 big thing → why it matters → 2-4 key numbers (each vs plan/baseline) → risk → what's next
Memo              Recommendation → stakes → strongest evidence → risks / counterarguments → decision needed (who, by when) → appendix
Request           The ask + deadline → why now → minimum info to decide → attachment/link
Analysis          Answer (with its key uncertainty) → significance → evidence & counterevidence → implications
Long-form         Apply recursively: one thesis; each section opens with its own point and why it matters

Useful axioms (use only when they label a real function): Why it matters, The big picture, By the numbers, Yes, but, What's next, The bottom line, Go deeper.

Output

Return the rewrite first. Then, unless the user asked for the text only, add 1-3 short lines on the biggest changes, and list anything you flagged as missing or any caveat you chose to keep. Don't narrate every edit.

For worked before/after rewrites, see examples.md.

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.