CodeTree Skill for Clawdbot

Clawdbot skill for querying code repositories using natural language. Built on vectorless RAG technology.

Installation

pip install codetree-skill

Usage

CLI Commands

# Index a repository
codetree-skill index /path/to/repo

# Query the codebase
codetree-skill query /path/to/repo "How does authentication work?"

# Show code structure
codetree-skill tree /path/to/repo

# Find symbol references
codetree-skill find /path/to/repo "UserService"

# Get statistics
codetree-skill stats /path/to/repo

As Clawdbot Skill

Copy the skill files to your Clawdbot skills directory:

# Find where the skill is installed
python -c "import codetree_skill; print(codetree_skill.__file__)"

# Copy to Clawdbot skills (adjust path as needed)
cp -r /path/to/codetree_skill ~/.clawdbot/skills/codetree/

Features

• 🚫 No Vector Database - Uses LLM reasoning instead of embeddings
• 🌳 AST Parsing - Understands actual code structure
• 🔗 Cross-File References - Tracks imports and dependencies
• 💬 Natural Language - Ask questions in plain English

Supported Languages

• Python
• JavaScript / TypeScript
• Go
• Rust
• Java

Requirements

• Python 3.10+
• OpenAI or Anthropic API key

Links

• CodeTree Main Package
• CodeTree MCP Server
• GitHub Repository