write
Skill1 starsChanged 3 months agoMITView on GitHub
Draft or revise a section of the document. Use when the user wants to write, draft, expand, revise, rewrite, improve, or continue work on any section such as the introduction, literature review, methodology, analysis, discussion, or conclusion. Also triggers on "start writing", "flesh out", "add content to", "continue working on".
40 other projects use a version of this.
Adopted by
Most recently taken up by
- apmantza/pi-lenstheir own version, first seen Nov 16, 2026
- j0yen/dotfilestheir own version, first seen Nov 16, 2026
- noelceta/content-enginetheir own version, first seen Nov 16, 2026
- waynesutton/convexskillstheir own version, first seen Nov 15, 2026
- waynesutton/builder-skillstheir own version, first seen Nov 15, 2026
- legibet/agent-skillstheir own version, first seen Nov 8, 2026
Used with
Projects with it also use sec-check, read and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
waynesuttonSkillsec-check
Security review checklist for Convex functions, auth logic, public queries, admin routes, webhooks, uploads, and AI-generated code. Use when reviewing code that touches user data, PII, or access control.
Also in 3 projects that use this
MicrosoftSkillread
Read pages from a document. Returns MinerU markdown (preferred) or PyPDF text, with optional page screenshots and sub-image fetching. When markdown is available, a figure_images_meta catalog is included — use the figures parameter to fetch specific sub-image pixels. Triggers include "read page 5", "what does page 12 say", "show me the chart on page 3", "extract pages 3-5".
Also in 6 projects that use this
ninehillsSkillhunt
Finds root cause before applying fixes for errors, crashes, regressions, failing tests, broken behavior, and screenshot-reported defects. Use when users report in any language errors, crashes, broken behavior, regressions, failing tests, screenshot evidence, or something that used to work and now fails. Not for code review or new features.
Also in 4 projects that use this
vltingSkillthink
Lightweight plan-to-execution pipeline. Explores codebase, runs council deliberation, presents plan for approval, then executes via /run.
Also in 7 projects that use this
waynesuttonSkillupdate-project-docs
Also in 3 projects that use this
genomewalkerSkillhealth
Soul system health check with remediation
Also in 5 projects that use this
Other versions40 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 40
- reactjs/react.dev★ 12ksame name
- tw93/claude-health★ 7.6ksame name
- tw93/Waza★ 7.6ksame name
- docker/docs★ 4.7ksame name
- waynesutton/markdown-site★ 628same name
- apmantza/pi-lens★ 530same name
- waynesutton/builder-skills★ 404same name
- waynesutton/convexskills★ 404same name
- yan-labs/yan-skills★ 307same name
- ohad6k/emulo★ 292same name
30 more not listed
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 repository13 other files it gives its agents.
- .claude/CLAUDE.mdCLAUDE.md
- factsSkill
- figureSkill
- full-reviewSkill
- ingestSkill
- onboardSkill
- plan-sectionSkill
- researchSkill
- reviewSkill
- save-materialSkill
- statusSkill
- teachSkill
- verify-sourcesSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 36 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 33
Every substantive claim carries a citation or rests on the document's own evidence.
- Very specific: line 34
Follow the section-specific structure your profile and structure.md require.
- Very specific: line 35
Prefer verified specifics to vague scale words.
- Distinctive: line 37
Keep the terminology fixed.
- Very specific: line 42
Run the pre-flight grep block from the bottom of .claude/rules/writing.md against the file you edited, and rewrite every hit.
“Common” is measured: kept by at least 6 in 10 of the 36 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
/write — Draft or revise a section
You are drafting part of a long document that has to hold together across months of work. Consistency with what already exists matters as much as the quality of the new prose.
1. Identify the target
Map the argument to a file in sections/. If the name is ambiguous, list the candidates and ask. If the section does not exist, check .claude/rules/structure.md before creating it: an unplanned section is usually a sign the argument has drifted.
2. Read before writing
In this order, and do not skip the first two:
docs/facts.md— every number, date, sample size, and terminology decision comes from here, not from memory and not from an older draft. Where a draft and this file disagree, this file wins..claude/rules/writing.md— style, citation format, and the banned-phrase list. These are hard constraints, not preferences..claude/rules/structure.md— this section's character budget and its place in the argument.docs/requirements.md— the institution's formal requirements and assessment criteria.- The target file, to see what is already there.
docs/plans/plan_<section>_*.md— an existing plan, if one was written.- Adjacent sections, for transitions and terminology.
docs/feedback/— feedback already given. Never regress a fix that was already made.- Relevant material: glob
docs/research/*andresources/**/*.mdand read what bears on this section.
3. Write
- Open with an orienting paragraph: what this section does and why it is here.
- Every substantive claim carries a citation or rests on the document's own evidence. A claim that has neither does not belong in the draft.
- Follow the section-specific structure your profile and
structure.mdrequire. - Prefer verified specifics to vague scale words. "4,312 responses over eleven weeks" beats "a large sample". If the exact figure is in
docs/facts.md, use it. - Tighten rather than inflate. If an explanation already exists in an earlier section, refer back to it in a clause instead of repeating it. Budget spent re-explaining is budget not spent on analysis.
- Keep the terminology fixed. One concept, one term, every time.
- Cite in the style set in
.claude/rules/writing.mdand nowhere else.
4. Self-check before reporting back
Run the pre-flight grep block from the bottom of .claude/rules/writing.md against the file you edited, and rewrite every hit. Then verify:
- Numbers, dates, and terms match
docs/facts.mdexactly. - Headings and labels name what is actually in them, rather than gesturing at a category.
- Any counted-list opener ("three reasons", "five categories") is followed by that many items.
- Forward references ("Section 5 discusses…") actually pay off in the target section. Read it and check.
- Citations added here exist in the reference list, or are flagged for adding.
5. Report
- Character count (
wc -m) against the section's budget. - What was cut, consolidated, or moved.
- New citations that need a reference entry.
- Grep hits found and fixed.
- Anything you could not source, stated plainly rather than papered over.
If the section is empty
Draft from the plan if one exists, otherwise from .claude/rules/structure.md and the available material. Say what you had to assume.
If the section has content
Revise. Ask what to focus on if the instruction is open-ended, and never silently rewrite prose that was already reviewed and approved.
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.

