
LlamaHub
by jbulger82
Local-first AI orchestrator and command center for llama.cpp with RAG, MCP tool integration, and browser automation.
What it does
LlamaHub serves as a comprehensive local command center for llama.cpp, transforming a raw LLM server into a powerful AI orchestrator. It connects local models to a rich ecosystem of MCP tools, providing a unified interface for complex tasks without requiring cloud dependency.
Tools
- Browser Automation: Integrated Playwright MCP for web interaction and scraping.
- Local RAG: Built-in document embedding and retrieval for persistent knowledge.
- System Utilities: Commands for disk usage, process management, and system status.
- Google Integrations: Gmail, Drive, and Calendar access via Google MCP.
- Deep Research: A multi-step research pipeline for complex topic exploration.
Installation
Add the LlamaHub proxy to your claude_desktop_config.json:
{
"mcpServers": {
"llamahub": {
"command": "node",
"args": ["/path/to/LlamaHub/proxy-server/index.js"]
}
}
}
Supported hosts
- claude
Quick install
git clone https://github.com/jbulger82/LLAMA_Hub && cd LLAMA_Hub && npm installInformation
- Pricing
- free
- Published
- 5/24/2026
- stars






