A more professional way to create crewAI projects using CLI and poetry

---

---

AI Nuggets

Install Poetry on Windows

1. Install pipx using pip
   • Open the command line and type:

     py -m pip install --user pipx  
     

   • You may receive a warning that the installation path is not on the system path.
   • Copy the given installation path.
   • Navigate to the directory of the installation path.
   • Inside that directory, run the command to ensure the path is added to the system environment:

     pipx ensurepath  
     

   • Verify the installation of pipx by checking its version:

     pipx --version  
     

2. Install Poetry using pipx
   • With pipx installed, you can now install Poetry by typing:

     pipx install poetry  
     

   • Check the installed version of Poetry:

     poetry --version  
     

Set Up a New Project with Crew AI

1. Create a new project directory
   • Navigate to your main project folder.
   • Create a directory for your project, for example, crew-ai-template.
   • Change into the new directory.
2. Launch Visual Studio Code
   • From inside the project directory, launch Visual Studio Code.
3. Create and activate a virtual environment
   • Create a virtual environment within Visual Studio Code.
   • Activate the virtual environment. The name of the virtual environment should appear before the prompt.
4. Install Crew AI packages
   • Install the main Crew AI package and the tools package by copying the command from the Crew AI documentation.
   • If there are issues, make adjustments as necessary to start the installation of the Crew AI package and dependencies.
   • Optionally, use pip freeze to list all installed packages and their versions in the requirements.txt file.
5. Use Crew AI CLI to create a project structure
   • Run the command:

     crew-ai create my-project  
     

   • This will create the structure of the Crew AI project.
6. Explore the project structure
   • There will be a README file with installation and running instructions.
   • The pyproject.toml file contains instructions for Poetry on setting up the virtual environment and installing needed packages.
   • The config folder contains configurations for agents and tasks in YAML format.
   • In crew.py, you can add custom tools or use tools provided by Crew AI.
   • main.py is where you run or “kick off” the project with input parameters.
7. Set up the project on a deployment server
   • Navigate to the commands section of the Poetry documentation to learn about options and settings.
   • To simulate a deployment, copy the project folder to a new location.
   • Open the project folder in Visual Studio Code.
   • Open the terminal and type:

     poetry lock  
     

   • This creates a virtual environment and a snapshot of the current packages in the lock file.
   • Next, install the packages by typing:

     poetry install  
     

   • To run the project, enter the command:

     poetry run my-project  
     

Conclusion

• Use Poetry and the Crew AI CLI command to create and set up Crew AI projects with a recommended structure.
• This method is professional and future-proof for setting up Crew AI projects.

Note: The transcript does not provide any URLs or specific CLI commands beyond what is listed above. If there were URLs mentioned in the video, they were not included in the transcript provided.