Aerostack
🧠
🔧 Developer Tools ✨ Featured

Knowledge Base + MCP

RAG knowledge base with MCP server built in

MCP Server RAG Vector Search AI LLM Pipeline Claude Desktop Compatible Auto-Discovery Manifest

About

A full RAG knowledge base with a built-in Model Context Protocol (MCP) server. Connect directly from Claude Desktop, Cursor, or any MCP-compatible client to search and chat with your documents.

Ingest docs via REST, chat via REST or MCP, and search via semantic retrieval — all from one deployment. The MCP manifest at `/.well-known/mcp.json` makes discovery automatic.

API Endpoints

POST /ingest
POST /chat
POST /search
GET /.well-known/mcp.json
POST /mcp
GET /health

How It Works

1

Document Ingestion

POST /ingest — documents chunked, embedded, and indexed in the vector database.

2

MCP Discovery

GET /.well-known/mcp.json — clients auto-discover available tools and resources.

3

Tool Execution

MCP clients call tools (search, chat) via JSON-RPC at POST /mcp.

4

RAG Response

Semantic retrieval + LLM generation returns grounded answers with sources.

Use Cases

🖥️

Claude Desktop Integration

Connect Claude Desktop to your private knowledge base via MCP for context-aware answers.

⌨️

Cursor AI Context

Give Cursor access to your internal docs so code suggestions are grounded in your architecture.

📚

API Documentation Hub

Ingest API docs and let any MCP client search and ask questions about your APIs.

🧑‍🤝‍🧑

Team Knowledge Portal

Centralize team knowledge and make it accessible from any MCP-compatible tool.

Quick Launch arrow_forward

Opens Aerostack dashboard to deploy this template

What's Included

check MCP Server
check RAG Vector Search
check AI LLM Pipeline
check Claude Desktop Compatible
check Auto-Discovery Manifest
check 6 API endpoints
check Edge deployed

Pipeline

database RAG — Vector search + retrieval
psychology LLM — AI text generation

Billing Model

metered

Pay per token used. Free tier included.

Tags

mcp knowledge-base rag claude cursor