Lspace MCP Server is an open-source API backend and MCP-compliant server that implements the Model Context Protocol (MCP). It enables developers to integrate intelligent knowledge base generation and management into workflows by connecting AI agents and other tools to managed content repositories. The project supports multi-repository management (local and GitHub), uses a dedicated entry point lspace-mcp-server.js, and operates with environment configuration (.env) and config.local.json to specify repositories and credentials (including GitHub PATs). The server can be run alongside or independently of a full API server (via npm run dev or npm start) and provides a dedicated MCP interface for programmatic interactions. It uses a dual-structure repository model: raw inputs stored under /.lspace/raw_inputs/ and a synthesized knowledge base at the repository root, with timeline tracking of document operations. LLM prompts are centralized in src/config/prompts.ts, enabling customization of document processing and knowledge-base generation. The README also documents how to configure client integrations (Cursor and Claude Desktop) to connect to the MCP server.