In short
Gemini agents connect to OmniChannels.ai through our remote MCP server at https://mcp.broadnet.me/mcp. In the Agent Development Kit, an MCPToolset uses Streamable HTTP with your token in an Authorization header; Gemini CLI takes the same URL and header in its settings. The agent can then send WhatsApp, SMS and RCS messages, use approved templates and read delivery reports. No Google marketplace app is involved.
What you can do
Customer updates from an ADK agent
An agent built with the Agent Development Kit sends shipping or appointment updates when your application asks.
Try asking
“Send a WhatsApp message to 971500000000 saying their order has shipped and will arrive tomorrow.”
Templates for Saudi and UAE customers
The agent lists approved WhatsApp templates and fills the Arabic or English version that fits.
Try asking
“List my approved Arabic WhatsApp templates and send order_update to 966500000000 with order number INV-400.”
Test sends from Gemini CLI
Developers send a test message or check a delivery report from the terminal while building.
Try asking
“Send a test SMS to my verified number and tell me the delivery status.”
Delivery analysis
A read-only agent with only the delivery report tool summarises failures for your team.
Try asking
“How many messages in yesterday's statements campaign failed, and what were the most common reasons?”
Combine with your business data
When the agent also has a tool for BigQuery, your CRM or your store, it can fetch recipients there and send through OmniChannels.ai.
Try asking
“Send a WhatsApp order update to customers whose orders shipped today.”
How it connects
OmniChannels.ai runs a remote MCP server over Streamable HTTP. Every request carries your API token in an Authorization: Bearer header, and each tool acts with the permissions, sender IDs and balance of that token, exactly as your own code would.
In the Agent Development Kit, add an MCPToolset whose Streamable HTTP connection parameters hold our URL and the Authorization header, and filter it to the tools the agent needs. An ADK agent can be deployed to Vertex AI Agent Engine; store the token in Secret Manager rather than in code. Gemini CLI reads MCP servers from its settings file, using httpUrl for the endpoint and headers for the token.
- Endpoint: https://mcp.broadnet.me/mcp (Streamable HTTP)
- Authentication: Authorization: Bearer <token>; use a temporary key while testing
- ADK: MCPToolset with Streamable HTTP connection parameters and headers
- Gemini CLI: mcpServers entry with httpUrl and headers
- Without MCP, the agent can call our REST API using the OpenAPI spec
Set it up
- 1
Request API access
Request test access and receive an API token. Test sends go only to your own verified numbers.
- 2
Get an MCP key
In the admin portal open Integrations, then MCP Integration, and copy a temporary key.
Guide: MCP server for AI agents - 3
Add the toolset to your agent
In ADK, create an MCPToolset with Streamable HTTP parameters for https://mcp.broadnet.me/mcp and the Authorization header, reading the token from an environment variable or secret.
- 4
Or add it to Gemini CLI
Add an mcpServers entry with httpUrl set to our endpoint and an Authorization header in headers, then restart the CLI.
- 5
Limit tools and confirm sends
Expose only the tools the agent needs and require a confirmation step before any send, since each send is a real, billed message.
- 6
Test, then deploy
Send a test message to your verified number and check the delivery report before deploying the agent to Vertex AI Agent Engine.
Vendor documentation
Google Gemini and Vertex AI is a trademark of its owner. OmniChannels.ai is not affiliated with or endorsed by Google Gemini and Vertex AI; this page describes a connection built with Google Gemini and Vertex AI's own MCP, HTTP or webhook features.
