notificationId
, which you can use to check the message status or fetch replies
Here’s a simplified example:
If you are not using the default provider configured in the Siren account, you can specify which provider to use by adding the
providerIntegration
field in the request. Please take a look at API Reference for the details.GET /message-status/{notificationId}
- Check whether the message was delivered, failed, or is still pending.GET /get-reply/{notificationId}
- Fetch replies to a message (currently supported for Slack only). This includes main messages and threaded responses from users. These endpoints help you build a full feedback loop between your system, the provider, and the user. For the complete details, refer to API Reference. You can also send notifications using MCP, AI Agent Toolkit, or SDK, depending on your system architecture.
Using MCP (Model Context Protocol)
If you are using MCP to send a notification quickly, follow these steps:1
Install or run the MCP server
You can either install it globally using
npm install -g @trysiren/mcp
, or run it directly with npx -y @trysiren/mcp
.2
Start the server with messaging tool enabled
Run the following command to enable the QuickSend tool:
npx -y @trysiren/mcp --tools=messaging.send --api-key=YOUR_SIREN_API_KEY
3
Configure your development environment
Connect your MCP-compatible environment (e.g., Claude, VS Code, Replit, Cursor) to the running MCP server.
Refer to the MCP documentation for environment-specific setup.
Refer to the MCP documentation for environment-specific setup.
4
Call the `messaging.send` tool
Once configured, use your assistant or client to trigger a notification using the
messaging.send
method — no workflow setup required.5
Track message status
Use tools like
messaging.getStatus
or the /message-status/{id}
endpoint to monitor delivery.Using AI Agent Toolkit
If you are using AI Agent toolkit to send a notification quickly, follow these steps:1
Install the Agent Toolkit
Install the SDK using your preferred package manager:
or
npm install @trysiren/agent-toolkit
or
yarn add @trysiren/agent-toolkit
2
Initialize the toolkit with your API key
Import
You can optionally configure the messaging tool permissions.
SirenToolkit
and pass your API key during initialization.You can optionally configure the messaging tool permissions.
3
Expose tools to your AI framework
Use the
getTools()
method to expose Siren tools (like send_message
) to frameworks like OpenAI, LangChain, or Vercel AI SDK.4
Trigger messages using `send_message` tool
The agent can now use function calls to send notifications on supported channel,— just like an API or MCP request.
5
Monitor delivery with `get_message_status` or logs
Use status-checking tools within the agent environment, or fallback to Siren’s tracking endpoints and logs.
Using SDK
If you are using the Siren SDK to send a notification quickly, follow these steps:1
Install the SDK
2
Initialize the client
3
Send a message
4
(Optional) Track delivery status
This example demonstrates usage with the Python SDK. Siren also provides a TypeScript SDK, and the same procedures can be followed using TypeScript as well. For details on the TypeScript SDK and complete SDK usage, refer to the SDK Documentation.