📄 中文摘要
ReviewRadar 是一个评论智能 MCP 服务器，旨在从大量应用评论中提取有价值的产品信号。该系统通过规则引擎和 LLM 路由对评论进行分类，存储嵌入以支持向量搜索，自动检测 P0/P1 问题，并通过 MCP 工具提供分析功能。其架构包括应用评论、规则引擎、LLM 路由器、向量数据库、分析和 MCP 工具，能够高效处理和分析用户反馈。
使用 TypeScript、规则、LLM 和向量搜索构建生产就绪的评论分析 MCP 服务器
📄 English Summary
Build a Production-Ready Review Analytics MCP Server with TypeScript, Rules, LLMs, and Vector Search
ReviewRadar is a review intelligence MCP server designed to extract valuable product signals from a large volume of app reviews. The system classifies reviews using a rules engine and LLM routing, stores embeddings for vector search, automatically detects P0/P1 issues, and exposes analytics through MCP tools. Its architecture includes app reviews, a rules engine, an LLM router, a vector database, analytics, and MCP tools, enabling efficient processing and analysis of user feedback.