self-review-docs
Skill0 starsChanged 54 days agoMITView on GitHub
Use when drafting or editing project prose — README, ADR, guide, runbook, changelog — so the result reads like a human wrote it for a reader arriving cold.
In this repository26 other files it gives its agents.
- ask-mattSkill
- code-reviewSkill
- create-prSkill
- draft-prSkill
- ready-prSkill
- developSkill
- implement-reviewSkill
- implementSkill
- implement-specSkill
- librarianSkill
- nemesisSkill
- refactorSkill
- research-codebaseSkill
- researchSkill
- review-docsSkill
- self-reviewSkill
- simplifySkill
- tddSkill
- test-auditSkill
- waterfall-challengeSkill
- waterfall-designSkill
- waterfall-implementSkill
- waterfall-planSkill
- waterfall-refactorSkill
- waterfall-reviewSkill
- zero-tech-debtSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
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 24
Prefer the Saxon word.
- Distinctive: line 28
Every claim carries its evidence or leaves.
- Distinctive: line 49
Use the project's own terms, taken from the artifact that defines them.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.
Files
Review the prose you just wrote, in the scope the user specified. Fix what you find and touch nothing outside it. Code comments and docstrings are covered elsewhere; this skill is for prose files.
Word choice
Headings and sentences are the same material. Apply Orwell's rules to both:
Never use a long word where a short one will do.
If it is possible to cut a word out, always cut it out.
Never use the passive where you can use the active.
Never use a foreign phrase, a scientific word, or a jargon word if you can think of an everyday English equivalent.
Latinate vocabulary sounds abstract; Anglo-Saxon words are short and physical. Prefer the Saxon word.
Claims
Every claim carries its evidence or leaves. "Significantly faster" needs a number. "Production-ready", "best-in-class", and "secure by default" need a definition nobody has written. A caveat the reader can act on beats a confident summary.
Voice
Match the documents around it: a runbook stays terse, a personal note stays personal, an engineering doc stays clear of marketing. Name the actor who does the work rather than hiding them behind the passive. Prefer the concrete noun to the abstraction.
Structure
Inverted pyramid. Lead with what the reader came for and push background beneath it. A section adds information its heading did not. When every paragraph opens on an abstract benefit and closes on a promise, the shape is generated rather than written.
Vocabulary
Use the project's own terms, taken from the artifact that defines them. Follow a reference before asserting what it says: a document that coins its own name for something the code already names will drift from it.
Standing alone
The reader arrives with no history of this change. A document that only makes sense to someone who watched it happen is overfitted — cut the narration of what this session discovered and keep what the reader needs. A note that names a migration in flight or an open question goes stale on its own schedule; put it where it will be found when it resolves, or leave it out.
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.

