GraphRAG Chat

Moderated multi-agent group chat.

• top-level agent routes questions to sub-agents by name or speciality
• top-level agent includes multi-memory delegation
• all memories saved to each memory implementation
• sub-agents can select which memory implementation to read from
• a Neo4j graph catalog is available for read/write to multiple knowledge graphs

Agents:

graphrag_chat_agent_v1 (root agent)
│
├── cypher_agent: direct read/write access to available knowledge graphs
│
├── agent_smith: Neo4j product specialist
│   └── (memory: in-memory service. TBD migrate to simple KG memory)
│
└── new_agent: Give them a fun name and topical expertise
      └── (memory: note the memory impl)

Developer Notes

1. Initialize the Python environment with uv

• Create a virtual environment:

  uv venv

• Activate the virtual environment:

  source .venv/bin/activate  # select the appropriate activate.* for your shell

• Install dependencies:

  uv sync 

2. Set up configuration files

• Copy the environment template and edit as needed:

  cp .env.template .env
  # Edit .env to set your environment variables

• Copy the Neo4j connections example and edit as needed:

  cp neo4j.example.json neo4j.json
  # Edit neo4j.json to configure your Neo4j connections

3. Run the agent using adk

• Start the agent web server:

  adk web

You should now be able to access the agent locally. See below for more details or troubleshooting steps.