writing-for-humans
Skill1 starsChanged 8 days agoMITView on GitHub
Copy-edit human-facing prose for clarity, preserving meaning, evidence, uncertainty, voice, and requested format. Leave code, data, quotations, and user-specified wording intact unless their editing is requested.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- Suge8/friedbuntheir own version, first seen Nov 8, 2026
- edgarcosta/skillztheir own version, first seen Nov 1, 2026
- jd-solanki/krutrimboxtheir own version, first seen Sep 27, 2026
- dork-labs/dorkostheir own version, first seen Sep 26, 2026
- timharris707/skillstheir own version, first seen Sep 5, 2026
- J03Fr0st/skillstheir own version, first seen Jul 24, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- timharris707/skills★ 44same name
- Suge8/friedbun★ 12same name
- dork-labs/dorkos★ 11same name
- edgarcosta/skillz★ 2same name
- jd-solanki/krutrimbox★ 1same name
- J03Fr0st/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 repository26 other files it gives its agents.
- codebase-designSkill
- code-reviewSkill
- deep-researchSkill
- diagnosing-bugsSkill
- domain-modelingSkill
- grillingSkill
- grill-with-docsSkill
- handoffSkill
- herdrSkill
- implementSkill
- orchestratorSkill
- prototypeSkill
- prSkill
- researchSkill
- retroSkill
- setup-matt-pocock-skillsSkill
- shoppingSkill
- tddSkill
- teachSkill
- ticket-readinessSkill
- to-specSkill
- to-ticketsSkill
- tutorSkill
- wayfinderSkill
- wizardSkill
- writing-for-agentsSkill
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
2
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
- Distinctive: line 16
Remove repetition, throat-clearing, filler, inflated phrasing, vague references, and unexplained shorthand.
- Distinctive: line 24
When maintaining or evaluating this skill, read MAINTENANCE.md for fixtures and model-specific hypotheses.
“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.
Files
Writing for humans
Copy-edit an existing draft so its reader can understand and use it. The user's request and source draft govern the edit; preserve claims, scope, evidence, uncertainty, citations, and requested voice. Use the named audience or infer a general informed reader while retaining domain precision.
Edit once, then compare
Make one editing pass through structure, sentences, and wording:
- Lead with the answer, outcome, or purpose. Give each paragraph one job and put qualifications beside the claims they limit. Keep a visible order; use descriptive headings, lists, or tables when they reduce the reader's search effort.
- Make actors and actions clear. Prefer concrete verbs, familiar words, and a stable term for each concept. Retain necessary technical language; explain it from the supplied context when possible. Preserve passive voice when the actor is unknown, irrelevant, or clearest in that form.
- Remove repetition, throat-clearing, filler, inflated phrasing, vague references, and unexplained shorthand. Split overloaded sentences at natural boundaries, keeping longer sentences where they carry a necessary relationship or qualification. Preserve the detail and rhythm appropriate to the audience, requested voice, and format.
Compare every changed passage with the source draft before returning. Check facts, numbers, named entities, actors, causality, time, modality, qualifiers, attribution, citations, links, examples, and requested actions. Preserve quotations, code, commands, identifiers, data, legal or safety wording, and explicitly requested wording unless editing them was authorized. Check that headings still describe their sections and that reordering preserves dependencies and necessary conditions.
Correct or revert any unsupported change. Do not invent evidence, examples, conclusions, or transitions that imply a new relationship. If a safe copy edit cannot resolve a material audience, factual, or scope conflict, preserve the passage and briefly identify the limitation; ask only when the missing decision blocks the requested result.
Finish when the changed passages pass that comparison and the prose fits the requested length and shape. Readability scores, fixed sentence lengths, and active voice are not substitutes for meaning and audience fit. Return the edited prose in the requested format, otherwise the input's format. Add a short edit note only when useful and permitted; omit it for clean-copy requests. Keep the editing checklist out of the delivered prose.
When maintaining or evaluating this skill, read MAINTENANCE.md for fixtures and model-specific hypotheses. It is not needed for an ordinary copy edit.
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
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.

