write-docs
Skill2 starsChanged 2 months agoMITView on GitHub
Write or critique prose that ships with a codebase — READMEs, guides and runbooks, CLI help and error text, code comments, AGENTS.md/CLAUDE.md, commit messages, PR titles and descriptions, and changelog entries. Use when drafting or editing these, including commits and PRs written as part of another task. Do NOT use to review the code in a PR (use code-review), or for blog posts (blog-review), design docs or ADRs (grill-me), or skills (write-skill).
43 other projects use a version of this.
Adopted by
Most recently taken up by
- alchemy-run/alchemy-effecttheir own version, first seen Nov 9, 2026
- alchemy-run/alchemytheir own version, first seen Nov 8, 2026
- adhamhaithameid/write-skillstheir own version, first seen Nov 8, 2026
- dzhng/jevgreptheir own version, first seen Nov 7, 2026
- oberon-systems/enggraphtheir own version, first seen Nov 6, 2026
- dzhng/skillstheir own version, first seen Nov 4, 2026
Used with
Projects with it also use close-spec, design-with-images 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.
dzhngSkillclose-spec
When a spec under specs/<feature>/ is done shipping (a write-spec build, or any planned task that produced a spec), archive it to specs/done/ and rewrite it from a build-plan into a durable rationale — the why, the principles, the invariants — pointing back to the real code for the how. Use when implementation has landed and the plan no longer matches what shipped, or the user says a feature/spec is finished. Pairs with [write-spec](../write-spec/SKILL.md) (the plan this closes) and [review](../review/SKILL.md) (run before closing).
Also in 5 projects that use this
dzhngSkilldesign-with-images
Explore UI and visual design with image generation, then iterate the real implementation against the user-selected concept. Use when brainstorming visual directions, offering mockup variants, or implementing an approved generated design; do not stop at the mockups or the first code approximation.
Also in 5 projects that use this
transitive-bullshitSkillcompare-screenshots
Compare screenshots against the intended design, distinguishing approved references from historical baselines. Use for iterative implementation matching, before/after visual review, objective image telemetry, or checking a lone capture for flat, empty, or badly framed content.
Also in 5 projects that use this
LucasMatuszewskiSkillaudit-agents
Audit and maintain agent configuration files (.claude/agents, AGENTS.md, CLAUDE.md, agent-memory). Checks for duplicates, outdated info, naming conflicts, wrong technical facts, and files over 200 lines. Fixes all issues and commits.
Also in 5 projects that use this
wolverin0Skillaudit-decisions
Standardize remediation decisions, baseline-policy entries, and legal-review banners for accepted, deferred, or policy-blocked audit findings.
Also in 4 projects that use this
rhnfzlSkillexplore-unknowns
Walk the user through mapping the unknowns of a task before, during, and after they build it, and hand them a four-quadrant map (settled ground, open decisions, taste, and landmines) they keep. Reach for this eagerly whenever a request is ambiguous or underspecified, the codebase or domain is unfamiliar, the user will "know it when they see it", a reference implementation must be understood before porting, a build-queue or architecture-changing task is about to start, a mid-build deviation needs capturing, or a finished change needs buy-in or verified understanding before merge. Trigger even when the skill is not named: "do a blindspot pass", "find my unknown unknowns", "interview me about this change", "brainstorm the approach for a task I am about to build", "help me scope this", "what am I missing here", "plan this before I build", "I know nothing about X in this codebase". When a territory profile is present it reads memory and the graph first and feeds the map into the build pipeline; without one it degrades to a plain general map. Not for trivial or mechanical tasks with clear acceptance criteria, single-fact lookups, pure debugging of a known symptom, or pure code review, those have their own tools. Do not use it to run the build itself; the map is the deliverable and implementation remains a separate task, whether it has not started or is already underway.
Also in 5 projects that use this
- +6 more from dzhng
Other versions43 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 43
- tldraw/tldraw★ 51ksame name
- tabler/tabler★ 42ksame name
- videojs/video.js★ 40ksame name
- PrefectHQ/prefect★ 24ksame name
- comet-ml/opik★ 23ksame name
- coder/coder★ 17ksame name
- browseros-ai/BrowserOS★ 14ksame name
- Agenta-AI/agenta★ 4.9ksame name
- liveblocks/liveblocks★ 4.8ksame name
- dzhng/jevgrep★ 3.2ksame name
33 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 repository27 other files it gives its agents.
- agent-guardrailsSkill
- benchmark-checklistSkill
- blast-radiusSkill
- blog-reviewSkill
- code-reviewSkill
- create-verification-skillSkill
- datavizSkill
- grill-meSkill
- grok-repoSkill
- hammerspoonSkill
- html-lessonSkill
- karpathy-planningSkill
- mechanism-vizSkill
- obsidian-captureSkill
- obsidian-searchSkill
- paper-searchSkill
- readbackSkill
- repo-evalSkill
- route-skillSkill
- sherlockSkill
- show-meSkill
- sketch-conceptSkill
- skill-curatorSkill
- test-writingSkill
- tla-modelSkill
- web-recapSkill
- write-skillSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 3 are distinctive, and 2 are very specific to this project.
0
Common
Most of the 42 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 19
Not this skill: blog posts and essays (use blog-review), design docs, ADRs (use grill-me), internal service wikis, or authoring a skill …
- Very specific: line 23
Read the implementation, schema, tests, --help, or diff that establishes the facts.
- Distinctive: line 27
Check that every claim, command, caveat, and boundary still matches the source.
- Distinctive: line 41
Use evidence instead of quality adjectives.
- Distinctive: line 42
Use one name for one thing.
“Common” is measured: kept by at least 6 in 10 of the 42 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-docs
Pick the artifact, read that file, then write. Do not load more than one — the conventions differ and mixing them produces documents that serve nobody.
| Writing… | Read | Reader you are serving |
| A README for a public lib, CLI, or agent tool | READMES.md | someone deciding in 30 seconds whether to use this |
| A developer guide, tutorial, runbook, migration, or deprecation | GUIDES.md | someone trying to complete a task without guessing |
CLI flags, subcommands, --help, error text, output contracts | CLI.md | someone typing at a prompt, and the script that wraps them |
| Code comments, docstrings, inline rationale | COMMENTS.md | the maintainer reading this during an outage |
AGENTS.md / CLAUDE.md operating guides | AGENTSMD.md | an agent with no memory of your project |
| Commit messages, PR descriptions, changelog entries | PR.md | a maintainer running git log in two years; for a PR, also the reviewer who must act today |
Not this skill: blog posts and essays (use blog-review), design docs, ADRs (use grill-me), internal service wikis, or authoring a skill (use write-skill).
Workflow
- Read the implementation, schema, tests,
--help, or diff that establishes the facts. - Name the reader and the action this document must enable.
- Choose one canonical term for each repeated concept. Preserve exact API names and identifiers.
- Draft through the selected artifact lens.
- Check that every claim, command, caveat, and boundary still matches the source.
- Apply the clarity mode below, then run the checker. Treat heuristic findings as questions, not automatic edits.
Clarity modes
- Strict — procedures, safety rules, agent workflows, CLI help, and recovery instructions. Also any text whose reader cannot ask a follow-up question: tool and function descriptions, error messages, and output another agent or script parses. Put conditions before commands. Use imperative steps, one action per step, direct verbs, simple tenses, and sentences of at most 20 words.
- Natural — README narrative, explanations, comments, and PR rationale. Keep terminology and actors clear, but allow contractions, semicolons, sentence rhythm, and passive voice when the actor is unknown or irrelevant.
Apply modes by content block. A README quickstart can be strict while its motivation remains natural.
Shared rules
- Lead with the reader's goal and shortest path. Put provenance and history later or omit them.
- Be genuine. Write as a maintainer helping a reader, not as a product sheet imitating confidence.
- Use evidence instead of quality adjectives. Give the measurement, example, or boundary behind the claim.
- Use one name for one thing. Do not rotate synonyms unless they identify different concepts.
- Name the doer and the action within the first eight words of a sentence. An abstract noun as subject (
the configuration of,the handling of) hides them: make the doer the subject and the action the verb. - One idea per sentence. Open on what the reader already holds and end on the new name or value; this outranks active voice. A condition longer than ten words becomes its own step.
- Break noun stacks of four or more words: “calibration of the connection resistance,” not “connection resistance calibration procedure.”
- Say each fact once. Omit template sections that do not change the reader's next action.
- Cut words that announce or reassure:
This document describes,simply,obviously. Readers assume the opposite of an intensifier, andsimplyblames them when the step fails. - A hedge in a procedure is a missing condition: write
if X, do Y, notyou may need to do Y. In explanation, one hedge states uncertainty and two in a sentence hide it. - State unsupported cases and tradeoffs. Simpler wording must not erase precise technical meaning.
Check the draft
python3 scripts/prose_lint.py --mode strict <path>
python3 scripts/prose_lint.py --mode natural <path>Use - for stdin and --json for structured findings. Exit 1 means strict errors; review-only findings return 0. Add <!-- prose-lint-ignore --> to ignore one line. The checker cannot detect synonym rotation without project context, so check terminology manually. This is not an STE certification or a substitute for factual review. See CLARITY-REFERENCE.md for rationale and sources.
Rewriting an existing document
- Grep tests and CI for strings pinned from the document before you restructure it. If a test pins wording, make it assert the requirement (order, a link's presence) instead of the exact text. Name the changed assertion in the PR.
- Find copies derived from it: package README, docs site, translations. Update them or state that they are out of scope.
Reviewing rather than writing
Same lens, inverted: verify the artifact against its source, then report what the target reader cannot do, cannot trust, or must read twice. Lead with the defect and its location, not with praise.
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.

