Paparats is a local Model Context Protocol (MCP) server designed to provide advanced semantic code search and rich code context for AI coding assistants like Claude Code, Cursor, and Windsurf. It indexes multiple code repositories using a Qdrant vector database and integrates with Language Server Protocol (LSP) for intelligent symbol navigation, real-time file watching, and AST-aware code chunking. Paparats returns only the most relevant code snippets to AI agents, enabling more accurate responses and significant token savings. It supports backend, frontend, and infrastructure code in a unified search and operates fully local for privacy and speed, making it ideal for developers, support teams, contractor ramp-up, and AI agent automation.
Visit Paparats's official website for product details and getting started.