Skip to content

Latest commit

 

History

History
223 lines (146 loc) · 11.5 KB

README.md

File metadata and controls

223 lines (146 loc) · 11.5 KB
Raw

🧰 AI Agent Service Toolkit

build status codecov Python Version GitHub License Streamlit App

A full toolkit for running an AI agent service built with LangGraph, FastAPI and Streamlit.

It includes a LangGraph agent, a FastAPI service to serve it, a client to interact with the service, and a Streamlit app that uses the client to provide a chat interface. Data structures and settings are built with Pydantic.

This project offers a template for you to easily build and run your own agents using the LangGraph framework. It demonstrates a complete setup from agent definition to user interface, making it easier to get started with LangGraph-based projects by providing a full, robust toolkit.

🎥 Watch a video walkthrough of the repo and app

Overview

Try the app!

Quickstart

Run directly in python

# At least one LLM API key is required
echo 'OPENAI_API_KEY=your_openai_api_key' >> .env

# uv is recommended but "pip install ." also works
pip install uv
uv sync --frozen
# "uv sync" creates .venv automatically
source .venv/bin/activate
python src/run_service.py

# In another shell
source .venv/bin/activate
streamlit run src/streamlit_app.py

Run with docker

echo 'OPENAI_API_KEY=your_openai_api_key' >> .env
docker compose watch

Architecture Diagram

Key Features

  1. LangGraph Agent: A customizable agent built using the LangGraph framework.
  2. FastAPI Service: Serves the agent with both streaming and non-streaming endpoints.
  3. Advanced Streaming: A novel approach to support both token-based and message-based streaming.
  4. Content Moderation: Implements LlamaGuard for content moderation (requires Groq API key).
  5. Streamlit Interface: Provides a user-friendly chat interface for interacting with the agent.
  6. Multiple Agent Support: Run multiple agents in the service and call by URL path
  7. Asynchronous Design: Utilizes async/await for efficient handling of concurrent requests.
  8. Feedback Mechanism: Includes a star-based feedback system integrated with LangSmith.
  9. Dynamic Metadata: /info endpoint provides dynamically configured metadata about the service and available agents and models.
  10. Docker Support: Includes Dockerfiles and a docker compose file for easy development and deployment.
  11. Testing: Includes robust unit and integration tests for the full repo.

Key Files

The repository is structured as follows:

  • src/agents/: Defines several agents with different capabilities
  • src/schema/: Defines the protocol schema
  • src/core/: Core modules including LLM definition and settings
  • src/service/service.py: FastAPI service to serve the agents
  • src/client/client.py: Client to interact with the agent service
  • src/streamlit_app.py: Streamlit app providing a chat interface
  • tests/: Unit and integration tests

Why LangGraph?

AI agents are increasingly being built with more explicitly structured and tightly controlled Compound AI Systems, with careful attention to the cognitive architecture. At the time of this repo's creation, LangGraph seems like the most advanced open source framework for building such systems, with a high degree of control as well as support for features like concurrent execution, cycles in the graph, streaming results, built-in observability, and the rich ecosystem around LangChain.

I've spent a decent amount of time building with LangChain over the past year and experienced some of the commonly cited pain points. In building this out with LangGraph I found a few similar issues, but overall I like the direction and I'm happy with my choice to use it.

With that said, there are several other interesting projects in this space that are worth calling out, and I hope to spend more time building with them soon:

  • LlamaIndex Workflows and llama-agents: LlamaIndex Workflows launched the day I started working on this. I've generally really liked the experience building with LlamaIndex and this looks very promising.
  • DSPy: The DSPy optimizer and approach also seems super interesting and promising. But the creator has stated they aren't focusing on agents yet. I will probably experiment with building some of the specific nodes in more complex agents using DSPy in the future.
  • I know there are more springing up regularly, such as I recently came across Prefect ControlFlow.

Setup and Usage

  1. Clone the repository:

    git clone https://github.com/JoshuaC215/agent-service-toolkit.git
cd agent-service-toolkit

  2. Set up environment variables: Create a .env file in the root directory. At least one LLM API key or configuration is required. See the .env.example file for a full list of available environment variables, including a variety of model provider API keys, header-based authentication, LangSmith tracing, testing and development modes, and OpenWeatherMap API key.

  3. You can now run the agent service and the Streamlit app locally, either with Docker or just using Python. The Docker setup is recommended for simpler environment setup and immediate reloading of the services when you make changes to your code.

Docker Setup

This project includes a Docker setup for easy development and deployment. The compose.yaml file defines two services: agent_service and streamlit_app. The Dockerfile for each is in their respective directories.

For local development, we recommend using docker compose watch. This feature allows for a smoother development experience by automatically updating your containers when changes are detected in your source code.

  1. Make sure you have Docker and Docker Compose (>=2.23.0) installed on your system.

  2. Build and launch the services in watch mode:

    docker compose watch

  3. The services will now automatically update when you make changes to your code:

    • Changes in the relevant python files and directories will trigger updates for the relevantservices.
    • NOTE: If you make changes to the pyproject.toml or uv.lock files, you will need to rebuild the services by running docker compose up --build.

  4. Access the Streamlit app by navigating to http://localhost:8501 in your web browser.

  5. The agent service API will be available at http://localhost:80. You can also use the OpenAPI docs at http://localhost:80/redoc.

  6. Use docker compose down to stop the services.

This setup allows you to develop and test your changes in real-time without manually restarting the services.

Local development without Docker

You can also run the agent service and the Streamlit app locally without Docker, just using a Python virtual environment.

  1. Create a virtual environment and install dependencies:

    pip install uv
uv sync --frozen
source .venv/bin/activate

  2. Run the FastAPI server:

    python src/run_service.py

  3. In a separate terminal, run the Streamlit app:

    streamlit run src/streamlit_app.py

  4. Open your browser and navigate to the URL provided by Streamlit (usually http://localhost:8501).

Development with LangGraph Studio

The agent supports LangGraph Studio, a new IDE for developing agents in LangGraph.

You can simply install LangGraph Studio, add your .env file to the root directory as described above, and then launch LangGraph studio pointed at the root directory. Customize langgraph.json as needed.

Contributing

Currently the tests need to be run using the local development without Docker setup. To run the tests for the agent service:

  1. Ensure you're in the project root directory and have activated your virtual environment.

  2. Install the development dependencies and pre-commit hooks:

    pip install uv
uv sync --frozen
pre-commit install

  3. Run the tests using pytest:

    pytest

Customization

To customize the agent for your own use case:

  1. Add your new agent to the src/agents directory. You can copy research_assistant.py or chatbot.py and modify it to change the agent's behavior and tools.
  2. Import and add your new agent to the agents dictionary in src/agents/agents.py. Your agent can be called by /<your_agent_name>/invoke or /<your_agent_name>/stream.
  3. Adjust the Streamlit interface in src/streamlit_app.py to match your agent's capabilities.

Building other apps on the AgentClient

The repo includes a generic src/client/client.AgentClient that can be used to interact with the agent service. This client is designed to be flexible and can be used to build other apps on top of the agent. It supports both synchronous and asynchronous invocations, and streaming and non-streaming requests.

See the src/run_client.py file for full examples of how to use the AgentClient. A quick example:

from client import AgentClient
client = AgentClient()

response = client.invoke("Tell me a brief joke?")
response.pretty_print()
# ================================== Ai Message ==================================
#
# A man walked into a library and asked the librarian, "Do you have any books on Pavlov's dogs and Schrödinger's cat?"
# The librarian replied, "It rings a bell, but I'm not sure if it's here or not."

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Roadmap

  • Get LlamaGuard working for content moderation (anyone know a reliable and fast hosted version?)
  • Add more sophisticated tools for the research assistant
  • Increase test coverage and add CI pipeline
  • Add support for multiple agents running on the same service, including non-chat agent
  • Service metadata endpoint /info and dynamic app configuration
  • More ideas? File an issue or create a discussion!

License

This project is licensed under the MIT License - see the LICENSE file for details.