RAG▌
by kalicyh
RAG offers cloud-based vector database, semantic search, and retrieval augmented generation with fast OpenAI-powered doc
Provides cloud-based document management and semantic search using OpenAI embeddings with in-memory vector storage, enabling retrieval-augmented generation workflows through document ingestion, metadata filtering, and cosine similarity search.
Both formats append explainx.ai attribution and the canonical URL for this MCP server listing.
best for
- / Building RAG applications with document Q&A
- / Local knowledge base search and retrieval
- / Document analysis with AI summarization
- / Prototyping semantic search features
capabilities
- / Upload and index documents with vector embeddings
- / Perform semantic search with cosine similarity
- / Generate AI summaries of retrieved content
- / Filter documents by metadata
- / Configure multiple embedding providers
- / Manage documents through web interface
what it does
A low-latency RAG (Retrieval-Augmented Generation) service that lets you upload documents and perform semantic search using OpenAI embeddings with local vector storage. Includes both direct retrieval and LLM-powered summary modes.
about
RAG is a community-built MCP server published by kalicyh that provides AI assistants with tools and capabilities via the Model Context Protocol. RAG offers cloud-based vector database, semantic search, and retrieval augmented generation with fast OpenAI-powered doc It is categorized under ai ml, analytics data.
how to install
You can install RAG in your AI client of choice. Use the install panel on this page to get one-click setup for Cursor, Claude Desktop, VS Code, and other MCP-compatible clients. This server runs locally on your machine via the stdio transport.
license
MIT
RAG is released under the MIT license. This is a permissive open-source license, meaning you can freely use, modify, and distribute the software.
readme
MCP-RAG: Low-Latency RAG Service
基于 MCP (Model Context Protocol) 协议的低延迟 RAG (Retrieval-Augmented Generation) 服务架构。
特性
- 极低延迟 (<100ms) 本地知识检索
- 双模式支持: Raw 模式 (直接检索) 和 Summary 模式 (检索+摘要)
- LLM 总结功能: 支持 Doubao、Ollama 等 LLM 提供商进行智能摘要
- 模块化架构: MCP Server 作为统一知识接口层
- 异步优化: 异步调用与模型预热机制
- 可扩展设计: 预留 reranker 与缓存模块接口
技术栈
- 后端框架: FastAPI
- 向量数据库: ChromaDB (本地部署)
- 嵌入模型: Doubao 嵌入 API (默认), 本地模型可选 (m3e-small / e5-small via sentence-transformers)
- LLM 模型: Doubao API, Ollama (本地部署)
- 协议: MCP (Model Context Protocol)
- 包管理: uv (现代化 Python 包管理器)
快速开始
1. 环境要求
- Python >= 3.13
- uv 包管理器
2. 安装依赖
# 基础安装 (仅云端API)
uv sync
# 如果需要使用本地embedding模型 (m3e-small, e5-small)
uv sync --extra local-embeddings
3. 启动服务
uv run mcp-rag serve
首次启动会报错(懒得改)
该命令同时启动 Streamable HTTP MCP 端点和管理界面,后续可以直接访问 HTTP 页面完成配置、上传与查询。
- 访问配置页面:
http://localhost:8060/config-page - 访问资料管理页面:
http://localhost:8060/documents-page - 访问 Swagger API 文档:
http://localhost:8060/docs
4. 配置管理
MCP-RAG 现在使用 JSON 文件进行持久化配置管理
data\config.json 文件存储配置信息,支持通过 Web 界面进行修改和保存。
默认配置示例:
{
"host": "0.0.0.0",
"port": 8060,
"http_port": 8060,
"debug": false,
"vector_db_type": "chroma",
"chroma_persist_directory": "./data/chroma",
"qdrant_url": "http://localhost:6333",
"embedding_provider": "zhipu",
"embedding_device": "cpu",
"embedding_cache_dir": null,
"provider_configs": {
"doubao": {
"base_url": "https://ark.cn-beijing.volces.com/api/v3",
"model": "doubao-embedding-text-240715",
"api_key": null
},
"zhipu": {
"base_url": "https://open.bigmodel.cn/api/paas/v4",
"model": "embedding-3",
"api_key": null
}
},
"llm_provider": "doubao",
"llm_model": "doubao-seed-1.6-250615",
"llm_base_url": "https://ark.cn-beijing.volces.com/api/v3",
"llm_api_key": null,
"enable_llm_summary": false,
"enable_thinking": true,
"max_retrieval_results": 5,
"similarity_threshold": 0.7,
"enable_reranker": false,
"enable_cache": false
}
注意:
- 仅测试豆包与智谱的向量模型,其他模型未测试
- 豆包的向量模型好像要下线了,不推荐使用豆包的向量模型
MCP 服务器配置
小智go服务端能通过 MCP 协议与 MCP-RAG 进行交互。以下是一个示例配置:
{
"mcpServers": {
"RAG": {
"url": "http://127.0.0.1:8060/mcp"
}
}
}
5. 使用 MCP 工具
{
"name": "rag_ask",
"arguments": {
"query": "查询内容",
"mode": "raw",
"limit": 5
}
}
许可证
MIT License
贡献
欢迎提交 Issue 和 Pull Request!
FAQ
- What is the RAG MCP server?
- RAG is a Model Context Protocol (MCP) server profile on explainx.ai. MCP lets AI hosts (e.g. Claude Desktop, Cursor) call tools and resources through a standard interface; this page summarizes categories, install hints, and community ratings.
- How do MCP servers relate to agent skills?
- Skills are reusable instruction packages (often SKILL.md); MCP servers expose live capabilities. Teams frequently combine both—skills for workflows, MCP for APIs and data. See explainx.ai/skills and explainx.ai/mcp-servers for parallel directories.
- How are reviews shown for RAG?
- This profile displays 68 aggregated ratings (sample rows for discoverability plus signed-in user reviews). Average score is about 4.6 out of 5—verify behavior in your own environment before production use.
Use Cases▌
Extended AI Capabilities
Add new capabilities to Claude beyond text generation
Example
Access external data sources, execute code, interact with tools and services
Transform Claude from chatbot to action-taking agent
Context Enhancement
Provide Claude with access to relevant context and data
Example
Load project documentation, access knowledge bases, query databases
Get more accurate, context-aware responses
Workflow Automation
Automate multi-step workflows combining AI and external tools
Example
Research → Summarize → Create document → Send notification
Complete complex tasks end-to-end without manual steps
Implementation Guide▌
Prerequisites
- ›Claude Desktop 0.7.0+ or Cursor IDE with MCP support
- ›Basic understanding of MCP architecture and capabilities
- ›Access credentials for integrated services (if required)
- ›Willingness to experiment and iterate on configuration
Time Estimate
15-60 minutes depending on server complexity
Installation Steps
- 1.Install MCP server: npm install -g [package-name] or via GitHub
- 2.Add server configuration to ~/.claude/mcp.json
- 3.Provide required credentials and configuration
- 4.Restart Claude Desktop to load new server
- 5.Test basic functionality with simple prompts
- 6.Explore capabilities and experiment with use cases
- 7.Document successful patterns for reuse
Troubleshooting
- ⚠MCP server not loading: Check config syntax, verify installation
- ⚠Connection errors: Check network, firewall, credentials
- ⚠Feature not working: Read server docs, check required parameters
- ⚠Performance issues: Monitor resource usage, check for network latency
- ⚠Conflicts with other servers: Check port assignments, namespace collisions
Best Practices▌
✓ Do
- +Read server documentation thoroughly before setup
- +Start with simple use cases to validate functionality
- +Test in non-production environment first
- +Monitor resource usage and performance
- +Keep servers updated for bug fixes and new features
- +Document configuration for team members
- +Use environment variables for sensitive configuration
✗ Don't
- −Don't grant overly permissive access to MCP servers
- −Don't skip reading security considerations in docs
- −Don't expose sensitive data without proper controls
- −Don't run untrusted MCP servers without code review
- −Don't ignore error messages—investigate root cause
💡 Pro Tips
- ★Combine multiple MCP servers for powerful workflows
- ★Create custom MCP servers for your specific needs
- ★Share successful configurations with team
- ★Use MCP inspector for debugging
- ★Join MCP community for tips and troubleshooting
Technical Details▌
Architecture
Model Context Protocol standardizes how AI hosts (Claude, Cursor) communicate with external tools and data sources through server implementations.
Protocols
- Model Context Protocol (MCP)
- JSON-RPC 2.0
- stdio or HTTP transport
Compatibility
- Claude Desktop
- Cursor IDE
- Custom MCP clients
When to Use This▌
✓ Use When
Use when you need Claude to access external data, execute actions, or integrate with tools. Best for extending AI capabilities beyond conversation.
✗ Avoid When
Avoid when native integrations exist (use official APIs directly), for real-time critical systems, or when security/compliance requires zero external dependencies.
Integration▌
- →Tool composition: Chain multiple MCP tools in workflows
- →Context augmentation: Provide AI with relevant external data
- →Action delegation: Let AI execute tasks on external systems
- →Bidirectional sync: Keep AI context and external systems in sync
Discussion
Product Hunt–style comments (not star reviews)- No comments yet — start the thread.
List & Promote Your MCP Server
Share your MCP server with the developer community
Ratings
4.6★★★★★68 reviews- ★★★★★Ganesh Mohane· Dec 28, 2024
We evaluated RAG against two servers with overlapping tools; this profile had the clearer scope statement.
- ★★★★★Mia Lopez· Dec 24, 2024
We wired RAG into a staging workspace; the listing’s GitHub and npm pointers saved time versus hunting across READMEs.
- ★★★★★Chinedu Diallo· Dec 20, 2024
Useful MCP listing: RAG is the kind of server we cite when onboarding engineers to host + tool permissions.
- ★★★★★Henry Menon· Dec 16, 2024
RAG is among the better-indexed MCP projects we tried; the explainx.ai summary tracks the official description.
- ★★★★★Layla Thompson· Dec 4, 2024
RAG reduced integration guesswork — categories and install configs on the listing matched the upstream repo.
- ★★★★★Kabir Park· Dec 4, 2024
We evaluated RAG against two servers with overlapping tools; this profile had the clearer scope statement.
- ★★★★★Arjun Srinivasan· Nov 23, 2024
I recommend RAG for teams standardizing on MCP; the explainx.ai page compares cleanly with sibling servers.
- ★★★★★Isabella Perez· Nov 23, 2024
RAG has been reliable for tool-calling workflows; the MCP profile page is a good permalink for internal docs.
- ★★★★★Sakshi Patil· Nov 19, 2024
RAG has been reliable for tool-calling workflows; the MCP profile page is a good permalink for internal docs.
- ★★★★★Kabir Choi· Nov 11, 2024
Strong directory entry: RAG surfaces stars and publisher context so we could sanity-check maintenance before adopting.
showing 1-10 of 68