orchid-rag-neo4j

Neo4j graph-store plugin for knowledge-graph RAG.

orchid-rag-neo4j is the Neo4j graph-store plugin for the Orchid AI framework. It provides Neo4jGraphStore backed by the Python Neo4j driver, auto-registering via Python entry points.

Installation

pip install orchid-ai orchid-rag-neo4j

What It Provides

Auto-Registration

[project.entry-points."orchid.graph_store_backends"]
neo4j = "orchid_rag_neo4j:_register"

Configuration

There is no orchid.yml key for the graph store — the framework does not construct one during bootstrap. Wire it from a startup hook (or embedded integration) before the graph is built:

# myproject/hooks/startup.py
from orchid_ai.rag.factory import build_graph_store


async def wire_graph_store(*, runtime, **_):
    runtime.graph_store = build_graph_store(
        graph_store_backend="neo4j",
        url="bolt://localhost:7687",
        username="neo4j",
        password="…",
    )

# orchid.yml
startup:
  hook: myproject.hooks.startup.wire_graph_store

The builder forwards its keyword arguments to Neo4jGraphStore(url, username, password).