agentleFS
Sign inStart free
paulo-yamagishipaulo-yamagishi/claude-code-setup/CLAUDE.md

claude-code-setup

CLAUDE.md2 starsChanged 9 months agoMITView on GitHub

This repository's primary artifact is skill/SKILL.md — a /setup skill that audits Claude Code configuration and suggests project-tailored improvements. The core/ directory serves as reference material for maintaining the skill's accuracy. BOOTSTRAP.md explains the /bootstrap skill for greenfield project setup. This is a documentation and skills repository — no application code, no build step, no tests. The primary artifacts are Markdown files that serve as Claude Code configuration guides and skills. No build, test, or lint commands — this is…

The file

We split this CLAUDE.md into the rules it gives an agent

9 rules. About 0 are common, 5 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No build, test, or lint commands — this is a pure Markdown repo.

    line 34
  2. Distinctive:

    JSON examples: Must be valid JSON — no comments, no trailing commas

    line 53
  3. Distinctive:

    File paths: Always correct relative paths from repo root

    line 54
  4. Very specific:

    Cross-references: Use relative paths like ../core/hooks-guide.md

    line 56
  5. Very specific:

    Release workflow: Run /release <major|minor|patch> to bump, update changelog, commit, and tag

    line 62
  6. Very specific:

    Tags: Every release is tagged vX.Y.Z

    line 63
  7. Distinctive:

    After completing a major task, evaluate:

    line 68
  8. Very specific:

    Create a new skill: if the task involved a repeatable workflow that would benefit from a dedicated /skill-name command

    line 69
  9. Distinctive:

    Update an existing skill: if you used a skill and found it missing steps, producing suboptimal results, or not covering edge cases you …

    line 70

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md 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
CLAUDE.md

CLAUDE.md — Meta Instructions

This repository's primary artifact is skill/SKILL.md — a /setup skill that audits Claude Code configuration and suggests project-tailored improvements. The core/ directory serves as reference material for maintaining the skill's accuracy. BOOTSTRAP.md explains the /bootstrap skill for greenfield project setup.

Structure

  • core/ — Language-agnostic reference guides (one topic per file)
  • projects/ — Language-specific extensions of core guides (Python, TypeScript, Go, Rust)
  • templates/ — Copy-paste ready config templates (agents/, skills/, rules/)
  • templates/bootstrap/ — Extracted templates used by /bootstrap skill
  • prompts/ — Prompt patterns and examples
  • skill/SKILL.md — The /setup skill (self-contained)
  • .claude/skills/bootstrap/ — The /bootstrap skill for new projects
  • .claude/skills/evolve/ — The /evolve skill for auto-maintenance
  • .claude/rules/ — Context-specific rules (auto-loaded by file path)

Architecture

This is a documentation and skills repository — no application code, no build step, no tests. The primary artifacts are Markdown files that serve as Claude Code configuration guides and skills.

Key Files

  • skill/SKILL.md — The /setup skill (also mirrored at .claude/skills/setup/SKILL.md)
  • .claude/skills/bootstrap/SKILL.md — The /bootstrap skill
  • .claude/skills/evolve/SKILL.md — The /evolve skill
  • skill/OUTPUT-FORMAT.md — Output template for /setup reports

Data Flow

  • /setup audits → emits setup-report block → /bootstrap consumes it
  • /evolve auto-maintains config after tasks (via Stop hook)
  • core/ guides are reference material for skill accuracy, not user-facing docs

Commands

No build, test, or lint commands — this is a pure Markdown repo. Useful commands:

# Validate docs
/verify-docs

# Audit this project's Claude Code config
/setup

# Apply suggested config
/bootstrap

# Auto-maintain Claude Code config after tasks
/evolve

Code Style

  • Markdown: GitHub-flavored, # for title, ## for sections, ### for subsections
  • JSON examples: Must be valid JSON — no comments, no trailing commas
  • File paths: Always correct relative paths from repo root
  • Line limits: Files under 300 lines where possible; skills under 200 lines
  • Cross-references: Use relative paths like ../core/hooks-guide.md

Versioning

This project uses Semantic Versioning. The current version lives in VERSION (single-line file). All notable changes are tracked in CHANGELOG.md following Keep a Changelog format.

  • Release workflow: Run /release <major|minor|patch> to bump, update changelog, commit, and tag
  • Tags: Every release is tagged vX.Y.Z
  • Bump guide: patch for fixes, minor for new skills/guides, major for breaking restructures

Skill Lifecycle

After completing a major task, evaluate:

  • Create a new skill if the task involved a repeatable workflow that would benefit from a dedicated /skill-name command
  • Update an existing skill if you used a skill and found it missing steps, producing suboptimal results, or not covering edge cases you encountered

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.

Check your setup, free