agentsclimarketplace

Humanize writing

Skill Brennan-Lewis/writers_voice/skills/humanize-writing

Revise prose so it sounds natural, specific, and appropriate to its writer while preserving meaning, facts, citations, formatting, and authorship. Use for humanizing, de-AI-ing, de-slopping, polishing, or auditing emails, posts, memos, documentation, reports, resumes, and other drafts for repetitive AI-style patterns.From its SKILL.md

Install
npx -y skills add Brennan-Lewis/writers_voice --skill humanize-writing

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 22 days oldThe repository was created 22 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.0 KB, 540 tokens by cl100k_base, as published. Nobody here has run it

Humanize Writing

Treat AI-style patterns as editing heuristics, not evidence of authorship and not a detector score to optimize.

Preserve the source

  • Preserve meaning, commitments, uncertainty, citations, names, numbers, and scope unless the user authorizes a substantive change.
  • Preserve grammatical person and ownership.
  • Never invent facts, sources, quotations, experiences, emotions, or first-person claims.
  • Keep valid domain terminology, code, commands, links, tables, metadata, and required templates.
  • Prefer a supplied voice sample over generic style rules.
  • Never promise that a rewrite is human-authored, undetectable, or able to pass an AI detector.

Run the edit

  1. Identify the audience, format, requested tone, and text to edit.
  2. Build a short fact ledger of names, dates, numbers, commitments, capabilities, causes, constraints, quotations, citations, and uncertainty.
  3. Diagnose clusters of formulaic patterns. Read references/ai-writing-patterns.md for a broad audit or when the first pass still feels generic.
  4. Rewrite conservatively:
    • Remove chatbot preambles, filler, inflated importance, vague attribution, forced symmetry, and generic conclusions.
    • Replace abstractions with concrete details already present in the source.
    • Prefer direct verbs and simple constructions when they are clearer.
    • Vary sentence and paragraph rhythm only as the writer and format permit.
    • Keep legitimate punctuation and compound modifiers; correct overuse rather than banning a punctuation mark.
  5. Compare the revision with the source and fact ledger. Restore anything that was dropped, strengthened, softened, or invented without authorization.
  6. Read the revision for canned phrasing, repetitive openings, abrupt transitions, and overly uniform rhythm.

Choose an editing mode

  • Plain language: Reduce jargon and sentence complexity without making the text casual.
  • Professional: Keep the prose concise, specific, and restrained.
  • Personal: Use opinions, humor, or informality only when they belong to the author or the user explicitly requests them.
  • Voice match: Follow the supplied sample without copying private facts, accidental errors, or irrelevant verbal tics.
  • Pattern audit: Identify style patterns without claiming AI wrote the text.

Return the result

Return the final rewrite first. Add a short note about material editorial choices only when it helps the user review the change. Show a detailed audit or draft-versus-final comparison only when requested.

When another skill follows this one, pass the complete rewrite and the fact ledger forward. Do not summarize the draft between stages.

What ships with it: 3 files

4.0 KB alongside SKILL.md

agents/

references/

Keep looking

Skills are one crate of 326,422. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.