weni-agents-toolkit 2.4.2a5

No project description provided

Weni Agents Toolkit

A Python library for creating and managing agent skills for the Weni platform. Build powerful conversational agents with type-safe components and a robust skill system.

Installation

pip install weni-agents-toolkit

Or with Poetry:

poetry add weni-agents-toolkit

Quick Start

Creating a Skill

import requests
from weni import Skill
from weni.context import Context
from weni.components import Text, QuickReplies
from weni.data import register_result

class GetAddress(Skill):

    def execute(self, context: Context):
        # This is how we would retrieve credentials or sensitive information from context
        token = context.credentials.get("X-App-Token")

        # This is how we would retrieve parameters for this agent tool
        cep = context.parameters.get("cep")

        # This is how we would retrieve global constants for this project
        api_url = context.globals.get("cep_api_url")

        # This block is the business logic for the sake of this example on retrieving an address based on the received CEP
        base_url = f"{api_url}/{cep}"
        response = requests.get(base_url, headers={"Authorization": f"Bearer {token}"})

        result = response.json()

        # This block is where data is registered for further analysis in the future
        register_result("address", result.get("street"))
        # This example I'm respoonding allowing quick replies message or a location
        return TextResponse(data=result)

Sending an Event

You can send custom events to the Weni Datalake using the Event class. This is useful for logging actions, integrations, or relevant data during skill execution.

from weni.events import Event

Event.register(Event(
    event_name="event_name",
    key="key_name",
    value_type="string",
    value="value",
    metadata={
        "agent_collaboration": {
            "agent_name": "agent_name",
            "input_text": "input_text"
        }
    }
))

Parameters:

• event_name: Event name.
• key: Unique key for the event.
• value_type: Value type (string, int, etc).
• value: Event value.
• metadata: (Optional) Additional event metadata.

Registered events are available for integration and further analysis.

Core Concepts

Context System

The context system provides secure access to:

context = Context(
    credentials={"api_key": "secret123"},     # Sensitive data
    parameters={"user_id": "123"},            # Skill parameters
    globals={"env": "production"}             # Global configuration
)

Available Components

• Text: Basic text messages
• QuickReplies: Interactive quick reply buttons
• ListMessage: Interactive list menus
• CTAMessage: Call-to-action messages
• Location: Location request messages
• OrderDetails: Order information messages
• Attachments: File attachments
• Header: Message headers
• Footer: Message footers

Response Types

• TextResponse: Simple text messages
• QuickReplyResponse: Messages with quick reply buttons
• ListMessageResponse: Interactive list menus
• CTAMessageResponse: Call-to-action messages
• LocationResponse: Location-based messages
• OrderDetailsResponse: Order information displays

Development

Prerequisites

• Python 3.9+
• Poetry

Setup

1. Clone the repository:

git clone https://github.com/weni-ai/agents-toolkit.git
cd agents-toolkit

2. Install dependencies:

poetry install

3. Run tests:

poetry run pytest

Code Quality

We use several tools to ensure code quality:

• pytest for testing
• mypy for type checking
• ruff for linting

Run all checks:

poetry run pytest
poetry run mypy weni
poetry run ruff check .

Contributing

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

License

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