This guide will walk you through using the Weni by VTEX CLI to create, deploy, and manage AI-powered agents, enhancing your customer service capabilities. By the end of this guide, you will be able to install the CLI, configure your project, and deploy your first agent.

With the CLI, you can:

• Create, deploy, and manage multiple AI agents
• Add custom tools to the agents
• Update agent configuration and behavior

To use the Weni by VTEX CLI, you must have: a Weni by VTEX Platform account; an account at weni.ai; and at least one project in your account.

The content is organized as follows:

• Installing the CLI
• Getting started with your project
• Troubleshooting
• Contributing to Weni by VTEX CLI

There are two installation methods:

• A quick installation with pip.
• A manual installation with Poetry, for development purposes.

To install the CLI using pip, open the terminal and run the following command:

To install the CLI manually, follow these steps:

1. Clone the repository by running the following command:

2. Install dependencies and make the CLI executable by running the following command:

To verify that Weni by VTEX CLI is installed, type the following command in your terminal:

If the installation was successful, your terminal should display something like this:

An error occurred while loading the image https://github.com/vtexdocs/dev-portal-content/main/docs/guides/Weni-by-VTEX/CLI/assets/weni-installation-verification.png

If you encounter any issues:

1. Check our GitHub Issues
2. Create a new issue with:
   • Your operating system
   • Python version (python --version)
   • Error message
   • Steps to reproduce

Step 1 - Log in to your Weni by VTEX account

1. Open the terminal and run the following command:

This should open your browser on the login page for authentication. If that does not happen, you can open the URL shown in the terminal.

2. Log in using your Weni by VTEX account and password.

Step 2 - List your projects

List your projects by running the following command:

This will show all projects you have access to and, next to each of them, its universally unique identifier (UUID).

Step 3 - Select a project

Select the project you want to work on by running the following command:

Replace your-project-uuid with the UUID from the project list.

Step 4 - Verify the current project

To verify you're working on the correct project, you can run the following command:

You will see the project's UUID in the terminal output.

Step 5 - Initialize a new agent

To create an agent, run the following command:

This command performs the following actions:

• Creates the necessary folder structure for your agent.
• Sets up a pre-built CEP (Complex Event Processing) tool.
• Creates an agent_definition.yaml file with the initial configuration. You can edit this file to customize your agent's behavior.

Understanding the source configuration

In the YAML above, there are two fields nested within the source field:

• path: This field specifies the directory containing your tool implementation.

  • The command automatically creates a tools directory and a folder called get_address nested within it.
  • In this path, you can find the files main.py, requirements.txt, and test_definition.yaml.

• entrypoint: This field informs the system which class to use.

  • Using main.GetAddress indicates that the system finds the file named main.py, uses the GetAddress class inside the file and that the class must inherit from the Tools base class.

Your project structure should look like this:

Step 6 - Implement the tool manually (if weni init fails)

If the command doesn't work, you can implement this tool manually following the next steps:

1. Create the tool directory

2. Create the file tools/get_address/main.py and insert the following:

• The class name GetAddress must match the class name in the entrypoint
• The file name main.py must match the file name in the entrypoint
• The class must inherit from Tool and implement the execute method

3. Create the requirements.txt file

Step 7 - Add credentials and global files (optional)

If you have credentials and global files, place these files in tools/get_address during local runs:

Step 7 - Deploy the agent

Deploy your agent by running the following command:

Custom Parameters

You can add more parameters to your tools:

Multiple Tools

Agents can have multiple tools:

This guide will help you get started with contributing to the project.

Prerequisites

• Python >= 3.12
• Poetry >= 1.8.5
• Git

Setting Up Development Environment

1. Fork the repository on GitHub

2. Clone your fork:

   
   

   _10

   git clone https://github.com/YOUR_USERNAME/weni-cli.git

   _10

   cd weni-cli

   
   

3. Install dependencies:

   
   

   _10

   poetry install

   
   

4. Create a new branch:

   
   

   _10

   git checkout -b feature/your-feature-name

   
   

Code Style

• Follow PEP 8 guidelines
• Use meaningful variable and function names
• Add docstrings to functions and classes
• Keep functions focused and concise

Testing

1. Write tests for new features:

   
   

   _10

   def test_your_feature():

   _10

   # Your test code here

   _10

   assert expected == actual

   
   

2. Run tests:

   
   

   _10

   poetry run pytest

   
   

3. Check coverage:

   
   

   _10

   poetry run pytest --cov

   
   

Documentation

1. Update documentation for new features
2. Add docstrings to new functions
3. Update README.md if needed
4. Add examples when relevant

Workflow

1. Create a feature branch
2. Make your changes
3. Write or update tests
4. Update documentation
5. Run tests locally
6. Commit your changes
7. Push to your fork
8. Create a Pull Request

Commit Messages

Follow conventional commits:

• feat: New features
• fix: Bug fixes
• docs: Documentation changes
• test: Adding or updating tests
• chore: Maintenance tasks

Example:

PR Guidelines

1. Create one PR per feature/fix
2. Include tests
3. Update documentation
4. Reference issues if applicable
5. Keep changes focused and minimal

PR Template

1. Automated checks must pass
2. Code review by maintainers
3. Documentation review
4. Changes requested if needed
5. Approval and merge

• Create an issue for bugs
• Ask questions in discussions
• Join our community channels

Recommended VSCode Extensions

• Python
• YAML
• markdownlint
• GitLens

Useful Commands

1. Version bump
2. Update CHANGELOG
3. Create release PR
4. Tag release
5. Deploy to PyPI

• Be respectful and inclusive
• Help others when possible
• Share knowledge
• Follow our Code of Conduct