Run private RAG and MCP skill workspaces with Yuxi
Deploy Yuxi when an agent team needs a private, multi-tenant knowledge workspace that combines document RAG, knowledge graphs, MCP tools, skills, sandboxes, and permissions.
npx skills add agentskillexchange/skills --skill run-private-rag-and-mcp-skill-workspaces-with-yuxi
Use Yuxi when an operator needs to stand up a private knowledge-agent workspace instead of sending documents and tool access through a hosted chat product. Agents can work over uploaded documents, knowledge graph context, MCP tools, skills, subagents, and sandboxed task execution while administrators keep data, model keys, tenants, and permissions under local control. The scope is operating a self-hosted RAG and agent workspace for repeatable knowledge workflows, not listing a generic framework or SDK.
What this skill actually does
Inputs and prerequisites: Docker Engine, Docker Compose, LLM API key, PostgreSQL, Redis, MinIO, Milvus, Neo4j.
Setup notes: Clone the repository at the documented release branch, run the initialization script to create environment settings and secrets, start the stack with docker compose up –build -d, confirm readiness at /api/system/ready, then open the local web app and initialize the administrator account.
Source and verification boundary: use https://xerrors.github.io/Yuxi/ as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.