agentleFS
henrymbuguakhenrymbuguak/claude-with-the-anthropic-api/.claude/skills/write-guide/SKILL.md

write-guide

Skill0 starsChanged 3 months agoMITView on GitHub

Use when: drafting one planned tutorial from guides/plan.json into docs/guides with measured facts and executable verification annotations.

6 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions6 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 6

    1. vercel/next.js★ 144ksame name
    2. readium/swift-toolkit★ 574same name
    3. karlo-ks/Infradeck★ 85same name
    4. runpod/runpod-plugins-official★ 50same name
    5. caiopizzol/tres-pinos★ 7same name
    6. sethsontech-max/claudeSkills★ 1same 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.

The file

Show its rules and text

We split this skill into the rules it gives an agent

18 rules. 0 are common, 13 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 13

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write exactly one guide identified by the guide ID in the request.

    line 9
  2. Very specific:

    Read only the selected entry in guides/plan.json plus its prerequisite entries.

    line 13
  3. Very specific:

    Read docs/contributing/writing-guides.md and docs/contributing/tutorial-template.md.

    line 15
  4. Very specific:

    Read docs/facts.json, every file in primary_code, and every file in tests from the selected entry.

    line 16
  5. Distinctive:

    Read prerequisite guide pages only when needed to avoid reteaching their concepts.

    line 17
  6. Distinctive:

    Stop without editing if the guide is not planned, is keyed, or its declared output already exists.

    line 19
  7. Distinctive:

    Never modify application code, tests, harness code, workflow files, dependencies, lockfiles, existing guides, or any other curriculum entry.

    line 30
  8. Distinctive:

    Follow the Google developer documentation style rules in the repository writing guide.

    line 34
  9. Very specific:

    Use the selected entry's descriptive output path.

    line 36
  10. Distinctive:

    Use four spaces for every paragraph, annotation, and code block that continues a numbered item.

    line 37
  11. Distinctive:

    Run every deterministic command before showing its output.

    line 38
  12. Distinctive:

    Use only offline commands.

    line 40
  13. Distinctive:

    Do not add a fact merely because it is interesting.

    line 42
  14. Distinctive:

    Add the guide to MkDocs in curriculum order.

    line 44
  15. Very specific:

    Keep a Next steps section.

    line 45
  16. Distinctive:

    Keep destructive build-from-empty steps manual.

    line 46
  17. Distinctive:

    Run these checks after writing:

    line 50
  18. Distinctive:

    Do not commit, push, open a pull request, or alter the generated branch.

    line 62

“Common” is measured: kept by at least 6 in 10 of the 6 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Write one planned guide

Write exactly one guide identified by the guide ID in the request. The repository's independent checks, not your judgment, decide whether the draft is valid.

Read before writing

  1. Read only the selected entry in guides/plan.json plus its prerequisite entries.
  2. Confirm the selected entry has status: "planned" and tier: "offline".
  3. Read docs/contributing/writing-guides.md and docs/contributing/tutorial-template.md.
  4. Read docs/facts.json, every file in primary_code, and every file in tests from the selected entry.
  5. Read prerequisite guide pages only when needed to avoid reteaching their concepts.

Stop without editing if the guide is not planned, is keyed, or its declared output already exists.

Allowed changes

Change only:

  • The selected entry's declared output file.
  • mkdocs.yml, adding one navigation entry for that output.
  • The selected entry's status, from planned to draft.
  • docs/facts.json only by running uv run python -m harness.measure_facts --root . --write docs/facts.json; never edit measured values manually.

Never modify application code, tests, harness code, workflow files, dependencies, lockfiles, existing guides, or any other curriculum entry.

Authoring rules

  • Follow the Google developer documentation style rules in the repository writing guide.
  • Teach only the concepts declared in the selected entry; the default budget is three.
  • Use the selected entry's descriptive output path. Never create a numbered guide-N.md filename.
  • Use four spaces for every paragraph, annotation, and code block that continues a numbered item.
  • Run every deterministic command before showing its output. Never invent output, versions, rankings, metrics, scores, errors, or API behavior.
  • Wrap every shell block with verify cmd or verify manual. A manual annotation requires a specific reason.
  • Use only offline commands. Do not access Anthropic, Voyage, or any other paid API.
  • Wrap every displayed scalar from docs/facts.json in a fact annotation and declare it in the selected plan entry.
  • Do not add a fact merely because it is interesting. Add only facts necessary to the lesson.
  • Link every declared prerequisite in Before you begin using its curriculum output path.
  • Add the guide to MkDocs in curriculum order. MkDocs previous/next navigation is the authoritative forward path.
  • Keep a Next steps section. Link only guides that already exist; mention planned guides without creating dead links.
  • Keep destructive build-from-empty steps manual. Never clear or overwrite source files in the checked-out repository.

Verification sequence

Run these checks after writing:

uv run python -m harness.verify_facts --root .
uv run python -m harness.verify_guide_flow --root .
uv run python -m harness.verify_guides "docs/guides/*.md"
uv run pytest -q
uv run ruff check .

The executable guide gate may reject a new command because it is not allowlisted. Do not weaken, bypass, or edit the allowlist. Leave the draft for a human to add a narrowly reviewed command rule.

Do not commit, push, open a pull request, or alter the generated branch. The workflow owns Git operations and changed-file enforcement.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.