This TypeScript-based MCP server implements the Model Context Protocol Server and exposes a single tool, chat, that relays questions to a configured AI chat provider. It is designed to integrate Claude Desktop workflows with any OpenAI SDK-compatible Chat Completion API (including OpenAI, Perplexity, Groq, xAI, PyroPrompts and more). Providers are configured via environment variables (AI_CHAT_KEY, AI_CHAT_NAME, AI_CHAT_MODEL, AI_CHAT_BASE_URL), enabling a single server instance to serve multiple backends by launching separate configurations. The README includes example Claude Desktop and LibreChat configurations showing how to point the MCP server to different backends. The server builds with npm, supports development with watch, and can be installed or run via npx or direct node invocation after building. For debugging, the MCP Inspector provides a browser-based debugging URL. You can install automatically via Smithery, or clone and build locally. The core capability is a lightweight chat tool that forwards user input to the chosen provider and returns responses through the MCP protocol, enabling cross-provider chat within Claude Desktop and LibreChat ecosystems.