manage-docs
Skill521 starsChanged 15 days agoMITView on GitHub
Create or update README files, AGENTS.md rules, and skills from actual code. Use before editing any README.md, AGENTS.md, or skill, or when condensing them.
We split this skill into the rules it gives an agent
12 rules. About 0 are common, 12 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈12
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 8
Write docs a developer or agent can follow without guessing.
- Distinctive: line 20
Read the target, its parent rules, and the code behind each claim.
- Distinctive: line 22
Ask one open question at a time: background, recommendation, then options.
- Distinctive: line 23
Write the approved version.
- Distinctive: line 25
Do not run deployment, secret upload, or build commands merely to document them.
- Distinctive: line 26
Do not commit or push unless requested.
- Distinctive: line 30
Write direct orders. Give each sentence one instruction.
- Distinctive: line 31
Keep lists to five items. Split longer lists by purpose.
- Distinctive: line 34
Remove filler, repeated summaries, and unverified claims.
- Distinctive: line 38
Check links, heading anchors, commands, and paths against the repository.
- Distinctive: line 39
Run the formatter check on edited files.
- Distinctive: line 40
Check the diff for unrelated edits, lost constraints, and exposed secrets.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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
Manage Docs
Write docs a developer or agent can follow without guessing.
Actions
| Target | Read |
README.md | readme.md |
AGENTS.md and CLAUDE.md | rules.md |
Skills in .agents/skills | skills.md |
Process
- Read the target, its parent rules, and the code behind each claim. Never print secret values.
- Show the proposed text before writing unless edits are authorized.
- Ask one open question at a time: background, recommendation, then options. Ask in plain text.
- Write the approved version. Do not ask again after approval.
Do not run deployment, secret upload, or build commands merely to document them. Do not commit or push unless requested.
Writing
- Write direct orders. Give each sentence one instruction. Put important rules first.
- Keep lists to five items. Split longer lists by purpose. Number steps when order matters.
- Copy commands, paths, and names from the code. State each command's working directory.
- Link to the owner of a rule or procedure. Never copy it.
- Remove filler, repeated summaries, and unverified claims.
Verify
- Check links, heading anchors, commands, and paths against the repository.
- Run the formatter check on edited files.
- Check the diff for unrelated edits, lost constraints, and exposed secrets.
- Report only checks actually run.
In this repository5 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- init-projectSkill
- manage-prsSkill
- review-codeSkill
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.