agentleFS
Flowerf19Flowerf19/agents-skills/architecture-docs/SKILL.md

architecture-docs

Skill3 starsChanged 4 days agoMITView on GitHub

Write and update project guidance using verified code and accepted decisions. Use when the user asks to write, audit, or simplify architecture docs, README.md, AGENTS.md, SKILL.md, or .agents guidance, or when an approved change needs related docs. Documentation only; do not edit runtime code.

4 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions4 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 4

    1. Seekers2001/docs-governance★ 320same name
    2. qshanx/docs-governance★ 253same name
    3. stempeck/agentfactory★ 3same name
    4. yiweiwan/skills★ 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.

  • In this repository5 other files it gives its agents.

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 14 are distinctive, and 0 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write guidance that is easy to follow and matches accepted decisions and verified facts.

    line 9
  2. Distinctive:

    Read the request, accepted decisions, and documents in scope.

    line 15
  3. Distinctive:

    Verify file paths, symbols, commands, and behavior against current source.

    line 16
  4. Distinctive:

    Update the existing document that owns the guidance.

    line 21
  5. Distinctive:

    Keep a root README focused on purpose, requirements, usage, checks, and important limits.

    line 25
  6. Distinctive:

    Follow the "Plain writing" section in AGENTS.md.

    line 29
  7. Distinctive:

    Use the requested language for project documents.

    line 31
  8. Distinctive:

    Use headings and numbered steps when order matters.

    line 33
  9. Distinctive:

    Keep project rules the agent cannot infer from the code.

    line 34
  10. Distinctive:

    When learning from other projects, use working code and tests as evidence.

    line 39
  11. Distinctive:

    Read the final text once as the reader.

    line 43
  12. Distinctive:

    Check links, paths, commands, names, versions, and contradictions across the changed documents.

    line 44
  13. Distinctive:

    Run safe checks needed to support factual claims.

    line 46
  14. Distinctive:

    Update affected existing project guides or module READMEs when the approved work changes their instructions.

    line 47

≈ “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

Architecture docs

Write guidance that is easy to follow and matches accepted decisions and verified facts.

Edit only requested documents or docs directly affected by an approved implementation. Do not edit code, tests, configuration, or dependencies to make a document true. Do not change plan approvals or task status unless the request includes them or an approved implementation has verified progress to record.

Step 1: Verify the facts

  1. Read the request, accepted decisions, and documents in scope. Follow relevant links in the existing project guides. Do not assume every project has the same documentation folders.
  2. Verify file paths, symbols, commands, and behavior against current source. Record the branch or commit when comparing projects.
  3. Separate current behavior, accepted work that is not implemented, and proposals. Report disagreements that affect the document.

Step 2: Choose where to write

  1. Update the existing document that owns the guidance.
  2. Change only the files needed for consistency. Link to shared rules instead of repeating them.
  3. For a new requested document, use the project's layout or a simple path for its purpose. Do not create a whole documentation tree.

Keep a root README focused on purpose, requirements, usage, checks, and important limits. Add a diagram only when it explains a verified design.

Step 3: Write clear instructions

Follow the "Plain writing" section in AGENTS.md. It applies ISO 24495-1: the reader must find, understand, and use the text on the first read.

  1. Use the requested language for project documents. Keep identifiers and commands exact.
  2. State the goal and scope first. Use short sentences, one idea per sentence, and a clear action.
  3. Use headings and numbered steps when order matters. Put optional or task-specific details under a clear condition.
  4. Keep project rules the agent cannot infer from the code. Remove repeated advice, stale paths, unsupported claims, and size limits with no stated purpose.
  5. Preserve security and data-loss warnings in full.

For skills, check when each skill is selected, what it may change, what it returns, and where it stops.

When learning from other projects, use working code and tests as evidence. Reuse useful design principles; keep project-specific databases, agent stages, business rules, and themes out of shared requirements.

Step 4: Review and report

  1. Read the final text once as the reader. Check that the goal comes first, each term is explained, and every step sits in the right order.
  2. Check links, paths, commands, names, versions, and contradictions across the changed documents.
  3. For a prompt rewrite, walk through representative requests and check permissions, steps, outputs, and handoffs. Use independent review when required and available.
  4. Run safe checks needed to support factual claims. Do not describe a proposed flow as tested.
  5. Update affected existing project guides or module READMEs when the approved work changes their instructions.

Report changed documents, check results, and unresolved questions. Keep document errors separate from confirmed runtime bugs.

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.