python

Siren Agent Toolkit (Python)

The Siren Agent Toolkit provides a unified Python interface and agent tools for interacting with the Siren MCP (Model Context Protocol) platform. It enables messaging, template management, user management, workflow automation, and webhook configuration, with seamless integration into popular agent frameworks like LangChain, OpenAI, and CrewAI.

Features & Capabilities

Messaging

• Send messages via various channels (Email, SMS, WhatsApp, Slack, Teams, Discord, Line, etc.)
• Retrieve message status and replies
• Support for template-based and direct messaging

Templates

• List, create, update, delete, and publish notification templates
• Create and manage channel-specific templates
• Support for template variables and versioning

Users

• Add, update, and delete users
• Manage user attributes and contact information

Workflows

• Trigger workflows (single or bulk operations)
• Schedule workflows for future or recurring execution
• Pass custom data to workflow executions

Webhooks

• Configure webhooks for status updates
• Set up inbound message webhooks
• Optional webhook verification with secrets

📋 Requirements

• A Siren API key (get one from Siren Dashboard)

Installation

pip install siren-agent-toolkit

For local development:

# From the python/ directory
pip install -e .

Usage

Basic Example

from siren_agent_toolkit.api import SirenAPI

# Initialize with your API key
api = SirenAPI(api_key="YOUR_API_KEY")

# Send a simple email message
result = api.run("send_message", {
    "recipient_value": "user@example.com",
    "channel": "EMAIL",
    "subject": "Important Update",
    "body": "Hello from Siren! This is an important notification."
})
print(result)

Examples

Complete working examples are available in the examples/ directory:

• examples/langchain/main.py — Using Siren tools with LangChain
• examples/openai/main.py — Using Siren tools with OpenAI
• examples/crewai/main.py — Using Siren tools with CrewAI

Development

Configuration

The toolkit supports flexible configuration options:

from siren_agent_toolkit.api import SirenAPI

api = SirenAPI(
    api_key="YOUR_API_KEY",
    context={"env": "production"}  # Optional environment configuration
)

Building Locally

# From the python/ directory
pip install -e .
# Install development dependencies
pip install -r requirements.txt

Running Tests

pytest tests/

License

MIT