liteMCP Client

Example Use Cases

• 1

  Expose MCP server tools to an OpenAI Agent SDK agent so the agent can call server-side capabilities directly.

• 2

  Bind MCP tools into a LangChain Chat model and execute returned tool_calls programmatically.

• 3

  Use direct OpenAI API tool schemas and then validate and execute selected MCP tool calls (including human-in-the-loop confirmation).

Description

liteMCP provides a simple, flexible way to bind tools served by an MCP server to popular LLM runtimes (e.g., OpenAI Agent SDK, LangChain, direct OpenAI API). It emphasizes minimal dependencies and straightforward interfaces to speed adoption and prototyping. It also supports custom tool-call validation (e.g., human-in-the-loop confirmation) and a configurable MCP server schema for enabling/disabling servers and excluding tools.