home / mcp / gomezzz299 mcp server
Provides a modular multi-agent MCP server that routes user questions to specialized agents and an LLM, with a web UI.
Configuration
View docs{
"mcpServers": {
"gomezzz299-mcp": {
"url": "http://backend:8000/process"
}
}
}You run an MCP server that routes user questions to specialized agents or an LLM, enabling collaborative, context-aware responses through a web UI and a modular backend. This guide shows you how to use, install, and configure your MCP server for local or networked environments.
Interact with the MCP server using the web GUI to ask questions like about the current date, location, weather, or a technical topic. The system decides whether to answer directly or delegate to a relevant agent, and agents collaborate to produce accurate, location-aware responses.
Prerequisites: you need Docker and Docker Compose installed on your machine. You will run the backend API and the Streamlit GUI in containers.
1. Prepare a working directory for the MCP server.
2. Create the Streamlit secrets file to point the GUI at the backend.
# gui/.streamlit/secrets.toml
server_url = "http://backend:8000/process"3. Start the stack with Docker Compose. This will build and run both the backend API and the frontend GUI.
docker-compose up --build4. Open the graphical interface in your browser. The backend backend API is exposed at http://localhost:8000 and the GUI at http://localhost:8501.
Configuration overview: the system uses a FastAPI backend for routing to specialized agents (date, location, weather, and expert LLM) and a SQLite database for contextual information. The UI is a Streamlit app. Docker containers simplify deployment, and the system is designed to run on a local network or across hosts.
Access notes: ensure ports 8000 and 8501 are exposed if you intend to access the server from another machine. Update the GUI secrets to point to the correct backend URL for remote deployments, for example http://server-ip:8000/process.
Security considerations: protect backend endpoints and limit access to trusted networks. Use environment isolation in production and consider securing the GUI with authentication if exposing beyond a private LAN.
Agent that returns the current date and time.
Agent that detects city and country via IP.
Agent that provides weather information for the current location.
LLM-based expert agent for specialized guidance.
Base class that standardizes error handling and response formatting for all agents.
Wrapper that interfaces with the Ollama LLM model.