whois

What is whois?

whois is a minimal MCP (Model Context Protocol) agent and server that lets you ask an AI about a person’s identity. It is designed as a simple example for developers exploring MCP-based agent architectures.

How to use whois?

Install the dependencies with pip install -r requirements.txt, then set the required API keys in your environment (ANTHROPIC_API_KEY, OPENAI_API_KEY, GEMINI_API_KEY, and optionally LLAMA_API_URL). Start the server with uvicorn server:app --reload, then run the agent with python agent.py "<name>" <provider> (e.g., python agent.py "Albert Einstein" openai).

Key features of whois

• Supports multiple AI providers: Anthropic, OpenAI, Gemini, Llama (Ollama)
• Minimal MCP agent and server implementation
• Queries an AI by person name
• Uses environment variables for secure API key management
• Runs as a lightweight ASGI server via uvicorn

Use cases of whois

• Asking an AI for a short biography of any person
• Testing and prototyping MCP agent workflows
• Comparing answers about the same person from different AI providers
• Learning how to build a minimal MCP-based client–server setup

FAQ from whois

What API keys are needed?

You need ANTHROPIC_API_KEY, OPENAI_API_KEY, and GEMINI_API_KEY set in your environment. For Llama support, also set LLAMA_API_URL (defaults to http://localhost:11434/api/generate).

Which AI providers are supported?

The agent supports Anthropic, OpenAI, Gemini, and Llama (via Ollama). The provider is specified as the second argument when running agent.py (e.g., anthropic, openai, gemini, llama).

How do I run the server?

After installing dependencies, run uvicorn server:app --reload in the project directory.

How do I run the agent?

Execute python agent.py "<name>" <provider> where <name> is the person you want to ask about and <provider> is one of openai, anthropic, gemini, or llama.