Quick Start
Start the Application
# Activate virtual environment
source .venv/bin/activate
# Start API, Web UI, and MCP HTTP endpoints
python cli.py serve
Available Services: - Web UI: http://localhost:5173 - API: http://localhost:8000 - MCP Tools: http://localhost:8000/mcp
Alternative Server Modes:
# API + MCP only (no Web UI)
python cli.py serve --api
# MCP stdio mode only (for legacy integrations)
python cli.py serve --mcp
Your First Crawl
Via Web UI
- Navigate to the "Crawl" page
- Click "New Crawl"
- Enter documentation name and URL
- Click "Start Crawl"
Via CLI
Upload Documentation
Upload Local Files
# Upload single file
python cli.py upload /path/to/docs.md --name "My Docs"
# Upload directory
python cli.py upload ./docs-folder --name "Documentation"
Process GitHub Repository
# Clone and process entire repository
python cli.py upload-repo https://github.com/user/repo --name "Project Docs"
# Process specific folder
python cli.py upload-repo https://github.com/user/repo --path docs
# Private repository with token
export GITHUB_TOKEN=ghp_your_token
python cli.py upload-repo https://github.com/user/private-repo
Search for Code
Via Web UI
Use the search bar on the dashboard or search page.
Via CLI
Via MCP
MCP tools are automatically available when the API server is running:
# List available MCP tools
curl http://localhost:8000/mcp/tools
# Execute a tool directly
curl -X POST http://localhost:8000/mcp/execute/get_content \
-H "Content-Type: application/json" \
-d '{"library_id": "library-name", "query": "search term"}'
For AI Assistants: Connect directly to http://localhost:8000/mcp using Streamable HTTP transport (MCP 2025-03-26 spec).