agentleFS
vosslabvosslab/vosslab-skills/skills/documentation/arch-docs/SKILL.md

arch-docs

Skill5 starsChanged 21 days agoMITView on GitHub

Create or refresh `docs/CODE_ARCHITECTURE.md` and `docs/FILE_STRUCTURE.md` from repository evidence. Use for architecture or file-layout documentation. Use other documentation skills for README, installation, usage, or repo-wide updates.

The file

Show its rules and text

We split this skill into the rules it gives an agent

20 rules. About 0 are common, 14 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈14

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create or refresh architecture and file-structure documentation from the current repo state.

    line 10
  2. Distinctive:

    Read README.md and AGENTS.md.

    line 16
  3. Very specific:

    Use rg --files to get a fuller inventory when needed.

    line 18
  4. Very specific:

    Check common manifests (pyproject.toml, pip_requirements.txt, package.json, Cargo.toml, go.mod, Makefile) when present.

    line 20
  5. Very specific:

    Check .gitignore for generated artifacts and caches.

    line 23
  6. Distinctive:

    Run git status -suno to see uncommitted changes.

    line 24
  7. Distinctive:

    Keep accurate sections and restructure for clarity when needed.

    line 27
  8. Distinctive:

    If unclear, add a verification task in the "Known gaps" section.

    line 29
  9. Very specific:

    Follow docs/MARKDOWN_STYLE.md and docs/REPO_STYLE.md.

    line 31
  10. Distinctive:

    Use present tense, concrete paths, and relative links.

    line 32
  11. Distinctive:

    Prefer short, scannable bullets.

    line 42
  12. Distinctive:

    Only include commands verifiable from repo files.

    line 43
  13. Distinctive:

    Avoid non-ASCII characters unless explicitly required by existing docs.

    line 44
  14. Distinctive:

    Prefer ISO-8859-1 only when the repo already uses it.

    line 45
  15. Distinctive:

    When showing a directory tree, use ASCII only.

    line 66
  16. Distinctive:

    Not allowed: box-drawing characters such as U+251C, U+2500, U+2502, U+2514

    line 69
  17. Very specific:

    Leave README.md to readme-docs, the sole owner of README content.

    line 87
  18. Very specific:

    Update docs/CHANGELOG.md directly when this skill runs as a standalone task.

    line 90
  19. Distinctive:

    Prefer to trim and correct existing docs before rewriting from scratch.

    line 96
  20. Distinctive:

    If the repo lacks needed evidence, add a "Known gaps" bullet with a verification task.

    line 97

≈ “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.

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

Arch docs

Overview

Create or refresh architecture and file-structure documentation from the current repo state. Base statements on observable files and configs, and record unknowns as verification tasks.

Workflow

  1. Inspect the repo
    • Read README.md and AGENTS.md.
    • Scan layout with ls -1 and ls -1 ./docs/.
    • Use rg --files to get a fuller inventory when needed.
    • Identify languages, build tooling, entry points, and main executables.
    • Check common manifests (pyproject.toml, pip_requirements.txt, package.json, Cargo.toml, go.mod, Makefile) when present.
    • Locate core logic, scripts, and tests.
    • Check .gitignore for generated artifacts and caches.
    • Run git status -suno to see uncommitted changes.
  2. Update policy
    • Remove stale details.
    • Keep accurate sections and restructure for clarity when needed.
    • State facts only; do not invent details.
    • If unclear, add a verification task in the "Known gaps" section.
  3. Apply style and consistency
    • Follow docs/MARKDOWN_STYLE.md and docs/REPO_STYLE.md.
    • Use present tense, concrete paths, and relative links.
    • Link file and folder names to their repo paths when mentioned, so they are clickable for details. Markdown links are created using the syntax [link text](URL), where "link text" is the clickable text that appears in the document, and "URL" is the web address or file path the link points to. This allows users to navigate between different content easily. Use file-path link text so readers know the exact filename (good: [docs/MARKDOWN_STYLE.md](docs/MARKDOWN_STYLE.md), bad: [Style Guide for Markdown](docs/MARKDOWN_STYLE.md)). Only include a backticked path when the link text is not the path.
    • Prefer short, scannable bullets.
    • Only include commands verifiable from repo files.
    • Avoid non-ASCII characters unless explicitly required by existing docs.
    • Prefer ISO-8859-1 only when the repo already uses it.
  4. Write docs/CODE_ARCHITECTURE.md
    • Title: "Code architecture"
    • Sections to include when applicable:
      • Overview
      • Major components (what, where, key dependencies)
      • Data flow (primary use case end-to-end)
      • Testing and verification (only if verifiable)
      • Extension points (where to add code/modules/integrations)
      • Known gaps (verification tasks only)
  5. Write docs/FILE_STRUCTURE.md
    • Title: "File structure"
    • Sections to include when applicable:
      • Top-level layout (key dirs/files with one-line purpose)
      • Key subtrees (large directories that need clarification)
      • Generated artifacts (what/where/git ignored)
      • Documentation map (docs location and root docs)
      • Where to add new work (code/tests/docs/scripts/data)

Repository structure

When showing a directory tree, use ASCII only. Do not use box drawing characters.

  • Allowed: |, +-, `-, spaces
  • Not allowed: box-drawing characters such as U+251C, U+2500, U+2502, U+2514

Example (ASCII only):

site_docs/
+- index.md
+- biochemistry/
|  `- topic01/
|     `- index.md
`- genetics/
   `- topic01/
      `- index.md
mkdocs.yml
  1. Wrap up
    • Save both files.
    • Leave README.md to readme-docs, the sole owner of README content. It links docs/CODE_ARCHITECTURE.md and docs/FILE_STRUCTURE.md from README once they exist. This skill owns only the two architecture docs.
    • Update docs/CHANGELOG.md directly when this skill runs as a standalone task; under delegate-manager-to-subagents, dispatch a docs subagent to add the entry.
    • Summarize what changed and what could not be verified.
    • Note that docs-only changes do not require tests unless otherwise requested.

Notes

  • Prefer to trim and correct existing docs before rewriting from scratch.
  • If the repo lacks needed evidence, add a "Known gaps" bullet with a verification task.

Example requests

  • "Update the architecture and file structure docs for this repo."
  • "Create CODE_ARCHITECTURE.md and FILE_STRUCTURE.md based on current files."
  • "Refresh the docs/FILE_STRUCTURE.md map after adding new folders."

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.