agentleFS
Sign inStart free
VercelKnown publishervercel-labs/marketing-team-eve-template/agent/subagents/content-marketer/skills/content-editing/SKILL.md

content-editing

Skill459 starsChanged 9 days agoMITView on GitHub

Use before handing back any long-form draft: edit in separate passes for sense, voice, reader payoff, proof, and specificity rather than re-reading and tweaking.

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    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:

    Is it proven?: Every claim needs a number, a source, a named example, or an explicit hedge.

    line 12
  2. Very specific:

    Read references/editing-checklist.md for the pass-by-pass checks, and references/common-problems.md for the recurring failure patterns and …

    line 17

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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

Content editing

Edit in passes, one dimension at a time. Re-reading a draft looking for everything at once finds nothing; each pass below has a single question. After a pass, check you haven't broken an earlier one.

  1. Does it make sense? Find the sentences carrying two ideas, the pronouns with no clear referent, the jargon a reader outside the company wouldn't know, and the point buried under qualifications. Fix structure before wording.
  2. Does it sound like us? One voice throughout. Watch for the register slipping into corporate abstraction in the middle sections, which is usually where the writer got bored. Check against the brand context's voice, not your own defaults.
  3. Does the reader care? For each section, ask what the reader gets from it. A feature described without a consequence fails this pass: say what it lets them do or cut it. If the answer is "it's context", it's probably too long.
  4. Is it proven? Every claim needs a number, a source, a named example, or an explicit hedge. Mark the ones that have none. You are not verifying them here, you're finding them: flag what needs a source rather than softening it into vagueness.
  5. Is it specific? Replace the general with the concrete. "Significantly faster" becomes the measured figure; "many teams" becomes the count you can support; "leading companies" becomes names you have permission to use. If you can't get the specific, drop the claim rather than keeping the vague version.

Then the mechanical pass: cut hedges and filler, break up sentences over about 25 words, kill passive voice where an actor exists, make sure each paragraph opens on its own point, and check the opening earns the second sentence.

Read references/editing-checklist.md for the pass-by-pass checks, and references/common-problems.md for the recurring failure patterns and their fixes.

This skill is about structure and argument. writing-quality carries the word-level rules, the AI phrases to avoid, and the plain-English list; load both.

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.

Reviews can't be read right now.

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