Agent skill
claude-md-creator
Use when creating CLAUDE.md files, updating existing CLAUDE.md, validating CLAUDE.md structure, or auto-fixing CLAUDE.md issues. Load for setting up project instructions, global guidelines, local overrides, or modular rules. Handles global (~/.claude/CLAUDE.md), project (.claude/CLAUDE.md), local (CLAUDE.local.md), and rules (.claude/rules/*.md) with smart project detection and template generation.
Install this agent skill to your Project
npx add-skill https://github.com/ingpoc/SKILLS/tree/main/claude-md-creator
SKILL.md
CLAUDE.md Creator
Create, validate, and maintain CLAUDE.md files with intelligent project detection.
Workflow Decision Tree
User Request
│
┌───────────┴───────────┐
│ What's needed? │
└───────────┬───────────┘
┌──────────────┼──────────────┐
│ │ │
Create new Update existing Validate/Fix
│ │ │
┌────┴────┐ ┌────┴────┐ ┌───┴────┐
│ │ │ │ │ │
Which Detect Merge Auto-fix Auto-fix
type? context changes issues issues
│ │ │ │ │
Detect Project Ask Run Run
context type user fix fix
│ │ changes script script
└────┬────┘ │ │ │
│ │ │ │
Generate Generate Validate Validate
content content output output
│ │ │ │
└───────────┴────────┴─────────┘
│
Write to
file
Quick Start
| Task | Command |
|---|---|
| Create project CLAUDE.md | scripts/generate-claude-md.py --type project |
| Update existing file | scripts/update-claude-md.py <path> |
| Validate file | scripts/validate-claude-md.py <path> |
| Auto-fix issues | scripts/auto-fix-claude-md.py <path> |
| Detect context | scripts/detect-claude-type.py |
Step 1: Determine CLAUDE.md Type
Script: scripts/detect-claude-type.py
Detection Logic:
| Context | Path | Size Target | When to Use |
|---|---|---|---|
| Global | ~/.claude/CLAUDE.md |
50-150 lines | Personal preferences across all projects |
| Project | .claude/CLAUDE.md |
100-300 lines | Team instructions for this project |
| Local | CLAUDE.local.md |
<50 lines | Personal overrides for this project |
| Rules | .claude/rules/*.md |
20-100 each | Modular topics by subject |
Step 2: Detect Project Type
Script: scripts/detect-project.py
Scans for project markers to generate smart defaults:
| Marker | Language | Framework | Template Used |
|---|---|---|---|
package.json + "next" |
TypeScript | Next.js | nodejs.md |
package.json + "react" + "vite" |
TypeScript | Vite React | nodejs.md |
requirements.txt + "fastapi" |
Python | FastAPI | python.md |
requirements.txt + "django" |
Python | Django | python.md |
Cargo.toml |
Rust | - | rust.md |
go.mod |
Go | - | go.md |
| None detected | - | - | general.md |
Step 3: Generate CLAUDE.md
Script: scripts/generate-claude-md.py
Template Selection:
Base template (assets/*.template.md)
+
Language template (assets/framework-templates/*.md)
+
Project-specific data (detected)
=
Final CLAUDE.md
Step 4: Validate Structure
Script: scripts/validate-claude-md.py
Checks Performed:
| Category | Check | Error Level |
|---|---|---|
| Frontmatter | Valid YAML fence | ❌ Error |
| Frontmatter | Required fields | ❌ Error |
| Structure | Section headers | ⚠️ Warning |
| Best practices | Line count | ⚠️ Warning |
| Best practices | Table format | ⚠️ Warning |
| Content | Command validity | ⚠️ Warning |
| Content | Path references | ⚠️ Warning |
Step 5: Auto-Fix Issues
Script: scripts/auto-fix-claude-md.py
Auto-Fixes:
| Issue | Fix | Backup |
|---|---|---|
| Missing frontmatter | Add YAML fence | ✅ Yes |
| Empty sections | Remove or placeholder | ✅ Yes |
| Malformed tables | Convert to proper Markdown | ✅ Yes |
| Extra blank lines | Collapse to 1 line | No |
| Inconsistent headings | Normalize to H2/H3 | ✅ Yes |
| Missing commands | Add from project detection | ✅ Yes |
Run modes:
# Dry run
./auto-fix-claude-md.py --dry-run <path>
# Auto-fix all
./auto-fix-claude-md.py <path>
# Fix specific category
./auto-fix-claude-md.py --category structure <path>
Step 6: Update Existing
Script: scripts/update-claude-md.py
Merge Strategy:
- Read existing CLAUDE.md
- Detect project changes
- Ask user what to update
- Preserve custom sections
- Write updated file
Best Practices
| Principle | Target |
|---|---|
| Tables > Prose | Use tables for commands, configs |
| Specific commands | Extract real commands from package.json |
| Line targets | Global: 50-150, Project: 100-300, Local: <50, Rules: 20-100 |
| Progressive disclosure | Quick start → detailed → references |
Resources
scripts/
| Script | Purpose | When to Use |
|---|---|---|
detect-claude-type.py |
Determine CLAUDE.md type | Auto-detection |
detect-project.py |
Scan project markers | Before generation |
generate-claude-md.py |
Create from templates | New file creation |
validate-claude-md.py |
Check structure | After edits |
auto-fix-claude-md.py |
Fix issues | Validation fails |
update-claude-md.py |
Update existing | Project changes |
references/
| File | Load When |
|---|---|
best-practices.md |
Writing content |
validation-rules.md |
Understanding errors |
project-detection.md |
Extending detection |
examples/ |
Real-world patterns |
assets/
| File | Purpose |
|---|---|
global.template.md |
Personal preferences |
project.template.md |
Team instructions |
local.template.md |
Personal overrides |
rule.template.md |
Modular topics |
framework-templates/*.md |
Language/framework additions |
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
scroll-storyteller
Use when creating interactive scroll-based storytelling experiences with mouse-following spotlight effects, animated SVG art, and the Anthropic design language. Load for explainer pages, product showcases, visual narratives, or any content needing immersive scroll storytelling with organic shapes and smooth animations. Supports GSAP-powered or lightweight CSS-only animations.
enforcement
Use when implementing hooks that BLOCK invalid actions, creating quality gates for state transitions, or enforcing tested:true verification. Load when designing enforcement mechanisms. Uses exit code 2 to block, JSON permissionDecision:deny, or updatedInput modification. Rules are instructions; hooks are enforcement.
async-programming-skill
This skill provides async/await patterns and best practices for concurrent programming
postgresql-skill
This skill provides PostgreSQL-specific patterns for database design, optimization, and transaction management
token-efficient
Use when processing 50+ items, analyzing CSV/log files, executing code in sandbox, or searching for tools. Load for data processing tasks. Achieves 98%+ token savings via in-sandbox execution, progressive disclosure, and pagination. Supports heredocs for multi-line bash.
mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
Didn't find tool you were looking for?