关于
Rag-mcp is an intricate retrieval-augmented generation (RAG) system operating as a Model Context Protocol (MCP) server, designed to provide advanced querying capabilities against a textual knowledge base. It leverages PostgreSQL with the pgvector extension to store and manage text embeddings, which are generated via an OpenAI-compatible API. The system exposes distinct search modalities, including semantic, question/answer, and style-based search, enabling AI agents and clients to perform nuanced information retrieval for RAG tasks. Its architecture is built for extensibility, allowing for the integration of custom search functionalities.
主要功能
- 2 GitHub stars
- Extensible framework for adding custom search modalities
- Supports semantic, question/answer, and style-based search modalities
- Model Context Protocol (MCP) Server for RAG
- Integrates with OpenAI-compatible embedding APIs
- Utilizes PostgreSQL with pgvector for efficient vector embeddings storage and search
使用案例
- Serving as a knowledge base backend for AI agents and language models
- Implementing custom retrieval-augmented generation (RAG) pipelines
- Enabling nuanced text search based on semantic meaning, Q&A, or stylistic similarity
Advertisement