AgentSkillsCN

init

通过分析当前代码库结构，对 CLAUDE.md 进行基础性引导与构建。

SKILL.md
--- frontmatter
name: init
description: Bootstrap CLAUDE.md by analyzing the current codebase structure
allowed-tools: Bash, Read, Glob, Grep, Write
disable-model-invocation: true

Initialize Project Context

Analyze the current codebase and generate a customized CLAUDE.md file.

Instructions

Step 1: Analyze Project Structure

bash
# Find project type indicators
ls -la
cat package.json 2>/dev/null | head -20
cat pyproject.toml 2>/dev/null | head -20
cat Cargo.toml 2>/dev/null | head -20
cat go.mod 2>/dev/null | head -20

# Find existing documentation
ls README* 2>/dev/null
ls docs/ 2>/dev/null

# Identify directory structure
find . -type d -maxdepth 2 | grep -v node_modules | grep -v .git | grep -v __pycache__ | grep -v .venv | head -30

# Find test directories
find . -type d -name "test*" -o -name "*tests" 2>/dev/null | head -10

# Check for existing Claude config
ls -la .claude/ 2>/dev/null
cat CLAUDE.md 2>/dev/null

Step 2: Detect Tech Stack

Identify:

  • •Language: Python, TypeScript, Go, Rust, etc.
  • •Framework: FastAPI, Express, React, etc.
  • •Build tools: npm, poetry, cargo, etc.
  • •Test runner: pytest, vitest, jest, etc.
  • •Linter: ruff, eslint, golint, etc.

Step 3: Extract Commands

Find available commands from:

  • •package.json scripts
  • •pyproject.toml scripts
  • •Makefile targets
  • •Justfile recipes
  • •Existing CI/CD configs

Step 4: Generate CLAUDE.md

Create a CLAUDE.md with:

markdown
# CLAUDE.md - Project Context for AI Agents

## Project Overview

**[Project Name]** is [detected description from README or package.json].

## Architecture Summary

- **[Component 1]**: [detected from directory structure]
- **[Component 2]**: [detected from directory structure]

## Quick Reference

### Required Environment
\`\`\`bash
[detected from .env.example or docs]
\`\`\`

### Common Commands
\`\`\`bash
# Lint
[detected lint command]

# Format
[detected format command]

# Test
[detected test command]

# Build
[detected build command]
\`\`\`

## Hierarchical Context

Claude Code auto-loads:
- `.claude/rules/*.md` - Project-wide rules
- `[component]/CLAUDE.md` - Component-specific context
- `knowledge/` - Domain expertise

## Testing Strategy

[detected from test structure]

## Acceptance Criteria

Before marking a task complete:
1. Code compiles/runs without errors
2. Tests pass
3. Linting passes
4. Committed and pushed

Step 5: Offer to Create Supporting Files

Ask if user wants to also create:

  • •.claude/rules/01-code-standards.md (based on detected stack)
  • •.claude/settings.json (with appropriate hooks)
  • •knowledge/ directory structure

Output

After generating CLAUDE.md, display:

markdown
## Project Initialized

**Detected Stack**: [language] + [framework]
**Test Runner**: [tool]
**Linter**: [tool]

### Files Created
- CLAUDE.md (project context)

### Recommended Next Steps
1. Review and customize CLAUDE.md
2. Add component-specific CLAUDE.md files
3. Create knowledge skills for your domain
4. Run `/prime` to load context