xerrors/Yuxi-Know
结合LightRAG 知识库的知识图谱智能体平台。 An agent platform that integrates a LightRAG knowledge base and knowledge graphs. Build with LangChain v1 + Vue + FastAPI, support DeepAgents、MinerU PDF、Neo4j 、MCP.
Python · ★ 645 · Fork 440 · 在 GitHub 查看
在手机上浏览源码、订阅每日 AI 趋势日报与项目架构深度解读:获取 CopoHub 客户端(iOS & HarmonyOS)→
GitHub Trending 上榜记录
xerrors/Yuxi-Know 已在 GitHub Trending 每日榜单上榜 1 天,最高排名第 11,首次上榜 2025-12-23,最近一次上榜 2025-12-23。
| 日期 | 排名 | Star | 当日增长 |
|---|---|---|---|
| 2025-12-23 | #11 | 52 | +52 |
同期上榜的项目
- rendercv/rendercv – CV/resume generator for academics and engineers, YAML to PDF
- exo-explore/exo – Run your own AI cluster at home with everyday devices 📱💻 🖥️⌚
- google/langextract – A Python library for extracting structured information from unstructured text using LLMs with precise source grounding and interactive visualization.
- yichuan-w/LEANN – RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate, and 100% private RAG application on your personal device.
- stan-smith/FossFLOW – Make beautiful isometric infrastructure diagrams
- vendure-ecommerce/vendure – The most customizable commerce platform built with TypeScript, NestJS and GraphQL.
Yuxi(语析)项目深度剖析与商业评估报告
Yuxi(语析)项目深度剖析与商业评估报告
1. 核心定位与商业落地价值
1.1 根本痛点:RAG 与 Agent 的"两张皮"困境
当前企业级 AI 落地面临一个结构性矛盾:RAG 系统与 Agent 系统各自为政。传统 RAG 平台(如 RAGFlow、Dify)擅长知识检索但缺乏任务执行能力;而 Agent 框架(如 LangGraph、CrewAI)擅长任务编排但知识来源不可控、无法溯源。Yuxi 的核心价值在于将 RAG 检索、知识图谱推理、多智能体编排、沙盒执行 统一到一个平台中,使 Agent 在运行时可实时检索知识库并获得带引用的推理结果,而非依赖模型参数内化的知识。
1.2 差异化定位:从 Demo 到团队级平台的跨越
多数开源 RAG/Agent 项目停留在单用户 Demo 阶段,Yuxi 明确瞄准 团队协作场景,其差异化护城河体现在:
多租户与权限治理:用户/部门两级权限模型,模型供应商统一配置,外部 API Key 集成,这是企业采购的硬性门槛。 知识图谱与 RAG 的深度融合:不仅做向量检索,还通过 Milvus + Neo4j 构建实体关系图谱,支持子图检索与推理。这是目前开源生态中极少见的完整实现(LightRAG 仅提供图谱构建,缺乏平台化封装)。 任务交付闭环:Sandbox 文件系统支持产物落盘、预览、下载,从"回答问题"升级为"交付任务成果"。
1.3 商业落地场景评估
在 README 中展示上榜徽章
[](https://github.fq6825.top/repo/xerrors/Yuxi-Know)