LangGraph RAG Research Agent Template

This is a starter project to help you get started with developing a RAG research agent using LangGraph in LangGraph Studio.

What it does

This project provides three runnable graphs implemented under src/graphs/:

• an "index" graph (src/graphs/index_graph.py)
• a "main" graph (src/graphs/main_graph.py)
• a "researcher" subgraph (part of the main graph) (src/graphs/researcher_graph.py)

The index graph takes in document objects and indexes them.

[{ "page_content": "RAG Research Agent是一种结合检索增强生成(RAG)技术与智能Agent能力的研究型智能体系统。它不仅具备基础的检索与生成能力，还融合了自我反思、工具链式调用及多智能体协同等高级特性，突破了传统大模型'仅会聊天、检索'的局限。" }]

Graph Relationships

The three graphs in this system have the following relationships:

1. Index Graph:

   • Independent graph for indexing documents into a vector store
   • Not directly connected to other graphs
   • Used to populate the vector store that other graphs query

2. Main Graph:

   • Main conversational agent
   • Contains the Researcher Graph as a subgraph
   • Uses the vector store populated by the Index Graph

3. Researcher Graph:

   • Subgraph of the Main Graph
   • Invoked by the 'conduct_research' node in the Main Graph
   • Generates queries and retrieves documents from the same vector store

Data Flow:

Index Graph --> Vector Store <-- Researcher Graph <-- Main Graph

Setup

Setup Vector Store

This template supports the following vector stores:

Elasticsearch (Local)

Elasticsearch is a distributed, RESTful search engine optimized for speed and relevance. For local development, we recommend using the official Docker image.

1. Pull the Elasticsearch Docker image:

   docker pull docker.elastic.co/elasticsearch/elasticsearch:9.1.4

2. Start Elasticsearch in a Docker container:

   docker run \
     -p 127.0.0.1:9200:9200 \
     -d \
     --name elasticsearch \
     -e ELASTIC_PASSWORD=your_password \
     -e "discovery.type=single-node" \
     -e "xpack.security.http.ssl.enabled=false" \
     -e "xpack.license.self_generated.type=trial" \
     docker.elastic.co/elasticsearch/elasticsearch:9.1.4

3. Set up your environment:

   • Create a .env file in your project root if you haven't already.
   • Add the Elasticsearch configuration to your .env file:

   ELASTICSEARCH_USER=elastic
   ELASTICSEARCH_PASSWORD=your_password
   ELASTICSEARCH_URL=http://localhost:9200
   

MongoDB Atlas

MongoDB Atlas is a fully-managed cloud database that includes vector search capabilities for AI-powered applications.

1. Create a free Atlas cluster:

• Go to the MongoDB Atlas website and sign up for a free account.
• After logging in, create a free cluster by following the on-screen instructions.

2. Create a vector search index

• Follow the instructions at the Mongo docs
• By default, we use the collection langgraph_retrieval_agent.default - create the index there
• Add an indexed filter for path user_id
• IMPORTANT: select Atlas Vector Search NOT Atlas Search when creating the index Your final JSON editor configuration should look something like the following:

{
  "fields": [
    {
      "numDimensions": 1024,
      "path": "embedding",
      "similarity": "cosine",
      "type": "vector"
    }
  ]
}

The exact numDimensions may differ if you select a different embedding model.

2. Set up your environment:

• In the Atlas dashboard, click on "Connect" for your cluster.
• Choose "Connect your application" and copy the provided connection string.
• Create a .env file in your project root if you haven't already.
• Add your MongoDB Atlas connection string to the .env file:

MONGODB_URI="mongodb+srv://username:password@your-cluster-url.mongodb.net/?retryWrites=true&w=majority&appName=your-cluster-name"

Replace username, password, your-cluster-url, and your-cluster-name with your actual credentials and cluster information.

Milvus

Milvus is a high-performance, open-source vector database designed for AI applications and similarity search.

1. Set up Milvus using Docker (for development) or use a managed Milvus service:

   • For Docker setup, follow the official Milvus installation guide
   • For Zilliz Cloud (managed Milvus service), sign up at Zilliz Cloud

2. Once you have your Milvus instance running, add the connection URI to your .env file:

MILVUS_URI=your_milvus_uri

For local development, this would typically be http://localhost:19530.

Setup Model

The defaults values for llm_model, embedding_model are shown below:

llm_model: ollama/qwen3:4b
embedding_model: ollama/bge-m3:latest

Follow the instructions below to get set up, or pick one of the additional options.

OpenAI

To use OpenAI's chat models:

1. Sign up for an OpenAI API key.
2. Once you have your API key, add it to your .env file:

OPENAI_API_KEY=your-api-key

Ollama (Local Models)

To use local models via Ollama:

1. Install Ollama
2. Pull the models you want to use:

   ollama pull qwen3:4b
   ollama pull bge-m3:latest

3. No API key is needed for local models.

Install dependencies

pip install -e .

For development, also install dev dependencies:

pip install -e ".[dev]"

Usage

Running examples

To run the examples in the template, use the following command:

python main.py

Visualizing graphs

To visualize the graphs in the template, run:

python src/visualize_graphs.py

This will display the graphs in your terminal and save them as PNG files in the graphs/ directory.

Running tests

To run the tests, use the following command:

make test

To run tests in watch mode:

make test_watch

To run tests with profiling:

make test_profile

Development

Formatting

To format the code, use:

make format

Linting

To lint the code, use:

make lint

Project Structure

.
├── src
│   ├── graphs
│   │   ├── __init__.py
│   │   ├── index_graph.py
│   │   ├── main_graph.py
│   │   └── researcher_graph.py
│   ├── shared
│   │   ├── __init__.py
│   │   ├── configuration_manager.py
│   │   ├── model_manager.py
│   │   ├── prompts.py
│   │   ├── retrieval.py
│   │   ├── retrieval_manager.py
│   │   ├── state.py
│   │   ├── text_encoder.py
│   │   └── utils.py
│   ├── __init__.py
│   ├── log_util.py
│   ├── sample_docs.json
│   └── visualize_graphs.py
├── tests
│   ├── integration_tests
│   │   ├── __init__.py
│   │   └── test_graph.py
│   ├── unit_tests
│   │   ├── __init__.py
│   │   └── test_configuration.py
│   └── __init__.py
├── Makefile
├── README.md
├── langgraph.json
├── main.py
└── pyproject.toml