Back to Apps

GraphRAG MCP Server
by rileylemm
Hybrid graph and vector retrieval system combining Neo4j and Qdrant for advanced semantic and relationship-based document search.
0 stars
Works in:Cursor
Exposes:ToolsResources
What it does
Connects LLMs to a hybrid retrieval architecture that merges the precision of vector search (via Qdrant) with the relational depth of graph databases (via Neo4j). This allows the AI to perform semantic queries while simultaneously expanding context through known relationships in the graph, overcoming the limitations of standard RAG.
Tools
search_documentation: Performs semantic search for information across the document store.hybrid_search: Combines vector similarity with graph-based relationship expansion for high-context retrieval.
Installation
Add the following to your claude_desktop_config.json:
{
"mcpServers": {
"GraphRAG": {
"command": "/path/to/graphrag_mcp/run_server.sh",
"args": []
}
}
}
Supported hosts
- Claude Desktop
- Cursor
Quick install
git clone https://github.com/rileylemm/graphrag_mcp.git && cd graphrag_mcp && uv install && uv run main.pyInformation
- Pricing
- free
- Published
- 8/5/2026
- stars
- 0
Categories
Choose your AI client and follow the steps below.
Cursor
Add to mcp.json in ~/.cursor/ or project .cursor/ folder.Claude Desktop
{
"mcpServers": {
"GraphRAG": {
"command": "/path/to/graphrag_mcp/run_server.sh",
"args": []
}
}
}





