Provides a flexible backend for AI models and custom tools, accessible via Model Context Protocol (MCP) and HTTP APIs.
This server offers a unique agentic architecture for integrating AI models and custom functionalities. It acts as a central hub where specialized agents encapsulate specific capabilities like OpenAI, Ollama, or file operations. These capabilities are exposed as tools, dynamically registered and routed through a central registry. The server supports dual access modes: the MCP protocol for seamless integration with Claude Desktop and standard HTTP endpoints for web and Streamlit UIs, enabling developers to build powerful AI applications with a modular and extensible design.