Ollama MCP Server

# Product Context ## Why This Project Exists Ollama MCP Server enables seamless integration of Ollama's local LLM capabilities into any MCP-powered application. It solves the problem of connecting local AI models to the broader MCP ecosystem, allowing developers to leverage local inference, privacy, and control. ## Problems It Solves - Lack of a standard interface for accessing Ollama from MCP tools. - Need for local, private LLM execution in MCP workflows. - Complexity of bridging CLI/HTTP APIs with MCP conventions. ## How It Should Work - MCP clients interact with the server using standard MCP tool calls. - Each Ollama operation (model management, execution, chat completion) is exposed as a tool. - The server translates MCP requests to Ollama CLI or API calls and returns results in MCP format. - Supports stdio transport (HTTP/SSE is a planned future enhancement). ## User Experience Goals - Simple integration for MCP clients. - Transparent access to all core Ollama features. - Fast, reliable, and robust error handling. - Configurable for advanced users (custom endpoints, timeouts, temperature).