# xerrors/Yuxi

可私有部署的多租户知识智能体平台：统一 RAG、知识图谱、多智能体、MCP/Skills、沙盒与权限管理。Self-hosted knowledge agent platform for RAG, knowledge graphs and multi-agent workflows.

Repository: https://github.com/xerrors/Yuxi
Canonical: https://ross.abutalabs.com/products/yuxi
Homepage: https://xerrors.github.io/Yuxi/
Language: Python
License: MIT
License Family: permissive
Topics: docker, kbqa, kgqa, llms, neo4j, rag, vue, fastapi, harness, agentic-rag, ai-agents, deepagents, document-ai, knowledge-base, knowledge-graph, langgraph, mcp, milvus, multi-agent, self-hosted
Last push: 2026-08-26T16:26:46+00:00

## Health v2 (maintenance only)
Score: 84/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 81, longevity 56
- inputs: {"age_days": 789, "days_push": 7, "days_rel": 47, "gap_med": 34, "n_releases_24m": 8}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6563, forks 977 (observed 2026-08-28T04:09:45.242133+00:00)

## What it is
Yuxi is a self-hosted, multi-tenant knowledge agent platform that unifies RAG, knowledge graph construction (Milvus + Neo4j), LangGraph multi-agent orchestration, MCP/Skills integration, sandboxed tool execution, and permission management in one workspace. It is deployed via Docker Compose with a Vue 3 frontend and FastAPI backend, and returns answers with traceable citations plus downloadable artifacts.

## Use cases
- build a private knowledge base Q&A chatbot with cited answers
- self-host a RAG platform for team documents
- extract entities into a knowledge graph and use it for retrieval
- orchestrate multi-step agents with MCP tools and skills
- manage multi-tenant knowledge access by user and department
- evaluate retrieval quality of a document knowledge base
- parse PDFs and Office docs into structured markdown for indexing

## When to choose
- you need full control over data, models, and permissions on your own infrastructure
- you want RAG plus knowledge graph reasoning in a single platform
- you need multi-tenant user/department isolation for knowledge bases and agents
- you want agents that execute tasks and deliver previewable file artifacts

## When to avoid
- you only need a simple single-user RAG script without multi-tenancy
- you cannot run the full Docker Compose stack (PostgreSQL, Redis, MinIO, Milvus, Neo4j)
- you need a fully stable release - the current version is still beta
- you want a managed SaaS rather than self-hosted deployment

## Facets
- artifact type: application
- maturity: active
- function: rag, agent-framework, mcp, chatbot, search-engine, web-framework, self-hosted, file-upload, pdf, workflow-automation
- domain: large-language-models, self-hosted, chatbots
- platform: self-hosted, python, windows
- tags: knowledge-base, knowledge-graph, multi-agent, multi-tenant, langgraph, milvus, neo4j, fastapi, vue, kbqa, kgqa, agentic-rag, document-parsing, sandbox, mcp-server, retrieval-augmented-generation, ai-agents, knowledge-graphs, natural-language-processing, docker, web-server, linux, macos

## Member repositories
- xerrors/Yuxi (main) score 84

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:45.242133+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:44:03.218423+00:00, confidence not recorded.
  - readme: https://github.com/xerrors/Yuxi (fetched 2026-08-28T04:09:45.242133+00:00, sha 94636653fbd7)
  - homepage: https://xerrors.github.io/Yuxi/ (fetched 2026-08-29T08:40:20.963595+00:00, sha fb678d8c230a)
- Data as of 2026-08-30T08:39:29.467469+00:00.