Skip to main content

CrewAI

CrewAI is a framework for orchestrating autonomous AI agents with specialized roles, tools, and goals.

When running a Crew, all of the agents run on the same device, usually using the same LLM. Using CrewAI together with Naptha allows you to run your CrewAI agents on different devices, while still being able to cooperate with each other.

Steps to run CrewAI agents on Naptha​

In this guide, we will show you how to automatically convert examples from the CrewAI examples repository to run on Naptha. In particular, we will show you how to run the Stock Analysis example.

You can checkout out the Naptha fork of the CrewAI examples repository to see:

  1. The decorated agent functions
  2. The agent module resulting from the conversion

Prerequisites​

  • Python >= 3.10, <= 3.13
  • Poetry package manager

1. Add Naptha SDK​

Since the CrewAI examples use poetry, we will need to add the Naptha SDK to the pyproject.toml file.

[tool.poetry.dependencies]
naptha-sdk = {git = "https://github.com/NapthaAI/naptha-sdk"}

2. Install dependencies:​

Execute via CLI and enter the environment:

poetry install
source .venv/bin/activate

3. Copy the .env.example file​

The CrewAI example has an .env.example file, so we will copy it to .env.

cp .env.example .env

And add the following variables to the .env file:

HUB_URL=wss://hub.naptha.ai/rpc
NODE_URL=https://node.naptha.ai
info

These variables are used to choose which Naptha Hub to connect to, and which Naptha Node to orchestrate the CrewAI agents.

4. Add your Naptha Hub credentials​

If you don't have a Naptha account, you can sign up for one by running the following command:

naptha signup

Or you can add your credentials manually to the .env file.

HUB_USERNAME=<your_naptha_hub_username>
HUB_PASSWORD=<your_naptha_hub_password>

5. Import Naptha SDK​

Put this in your main Python file: stock_analysis/crew.py:

from naptha_sdk.client.naptha import agent as naptha_agent

We recommend importing as naptha_agent to avoid naming conflicts.

6. Decorate​

Label the agent functions that you would like to run on different devices:

@agent
@naptha_agent("financial_agent")
def financial_agent(self) -> Agent:
    return Agent(
        ...
    )

...

@agent
@naptha_agent("research_analyst_agent")
def research_analyst_agent(self) -> Agent:
    return Agent(
        ...
    )

...

@agent
@naptha_agent("financial_analyst_agent")
def financial_analyst_agent(self) -> Agent:
    return Agent(
        ...
    )

...

@agent
@naptha_agent("investment_advisor_agent")
def investment_advisor_agent(self) -> Agent:
    return Agent(
        ...
    )

7. Convert (automatically)​

Process and package decorated functions:

python src/stock_analysis/crew.py

This creates a folder named agent_pkgs, which contains your "Napthafied" agent functions. The SDK translates your code into a format compatible with other agents on Naptha.

8. Test​

cd agent_pkgs/<agent_name>
poetry install
poetry run python <agent_name>/run.py

9. Publish​

Enter this command:

naptha publish

This command publishes all agents in the agent_pkgs folder to the Naptha Hub specified in your .env file.

10. Deploy agent​

naptha create <agent_name>

This command deploys the agent to the Naptha Node specified in your .env file.

11. Run deployed agent​

naptha run <agent_name>

Check for your expected output.