--- title: "n8n + OmniChannels.ai: messaging steps in your n8n workflows" description: "Use n8n with OmniChannels.ai: give an AI Agent our messaging tools via the MCP Client Tool node, or send and track messages with HTTP Request and Webhook nodes." canonical: "https://omnichannels.ai/integrations/n8n/" language: "en" updated: "2026-10-10" publisher: "BroadNet Technologies" brand: "OmniChannels.ai" --- # n8n + OmniChannels.ai: messaging steps in your n8n workflows Canonical: https://omnichannels.ai/integrations/n8n/ Language: en Publisher: OmniChannels.ai by BroadNet Technologies **In short:** n8n connects to OmniChannels.ai in two ways. An AI Agent node can use our messaging tools through the MCP Client Tool node pointed at https://mcp.broadnet.me/mcp with a Bearer token. Fixed workflows use an HTTP Request node to call the REST API and a Webhook node to receive delivery statuses and inbound messages. There is no dedicated n8n node; both routes use n8n's built-in nodes. Connect via MCP · Connect via API & webhooks ## What you can do ### Send an OTP or confirmation the same way every time When a form, app or database step produces a code or booking, an HTTP Request node sends it by SMS or WhatsApp through the REST API, with no AI involved. ### Send a WhatsApp template when a new row is added When a new row appears in a sheet or database that n8n watches, send an approved WhatsApp template with the row's values as its variables, to contacts who have opted in. ### Alert the team when a delivery fails A Webhook node receives delivery statuses; when one is FAILED or REJECTED, n8n posts an alert to your team chat or opens a ticket instead of resending automatically. ### Keep message status in your own records Store the message ID returned by each send, then update the matching record as SENT, DELIVERED and READ statuses arrive, ignoring late or duplicate events. ### Let an AI Agent answer delivery questions An AI Agent with read-only MCP tools looks up templates and delivery reports and answers your team's questions in a chat workflow. > Try asking: "Which messages to 966500000000 failed this week, and why?" ### Draft with AI, send after approval An AI Agent drafts an Arabic and English message, a team member approves it, and a fixed HTTP Request node sends the approved text. > Try asking: "Draft a short WhatsApp reminder in Arabic and English for tomorrow's appointments, for me to approve." ## How it connects For AI workflows, attach an MCP Client Tool node to an AI Agent node, set the endpoint to https://mcp.broadnet.me/mcp with Streamable HTTP transport, and store your key in a Bearer Auth credential. Choose only the tools the agent needs: n8n workflows usually run unattended, so nobody approves each tool call. For fixed workflows, an HTTP Request node calls the REST API at https://api.broadnet.me with a Bearer Auth credential. Single SMS takes a JSON body; bulk and most other channel sends take multipart form data, which the node supports. A Webhook node, whose production URL our team or your account admin sets as the webhook on your account, receives delivery statuses and inbound messages. - MCP: MCP Client Tool node, Streamable HTTP, Bearer Auth credential - API: HTTP Request node with JSON or form-data body - Events: Webhook node production URL configured on your account - Use fixed HTTP Request steps for OTPs and order notifications - Older n8n versions offer only SSE for MCP; update n8n if yours does ## Set it up 1. **Request API access:** Request test access and receive an API token. Test sends go only to your own verified numbers. (Guide: https://omnichannels.ai/developers/quickstart/) 2. **Store the token as a credential:** Create a Bearer Auth credential in n8n holding your API token or MCP key, so it stays out of exported workflows. (Guide: https://omnichannels.ai/developers/authentication/) 3. **Add the sending step:** For an AI Agent, attach an MCP Client Tool node with the endpoint https://mcp.broadnet.me/mcp and only the tools it needs. For a fixed workflow, add an HTTP Request node that POSTs to the channel's send path. (Guide: https://omnichannels.ai/developers/mcp-n8n/) 4. **Receive delivery statuses:** Add a Webhook node with method POST, activate the workflow, and send its production URL to our team or your account admin. (Guide: https://omnichannels.ai/developers/webhooks/) 5. **Test with your own number:** Run the workflow against your verified number and check that the send and its status events appear in the executions list. ## Example workflow Trigger in n8n (A new row, form entry, schedule or chat message) → n8n workflow (HTTP Request node, or AI Agent with MCP Client Tool) → OmniChannels.ai (REST API or MCP server sends the message) → Customer's phone (WhatsApp, SMS or RCS message delivered) → n8n Webhook node (Receives delivery statuses and inbound messages) ## Questions ### Is there an official OmniChannels.ai node for n8n? No. n8n connects through its built-in nodes: the MCP Client Tool node for AI Agents, and the HTTP Request and Webhook nodes for the REST API and delivery statuses. ### Should I use the AI Agent or the HTTP Request node? Use the HTTP Request node for anything that must send the same way every time, such as OTPs and order notifications. Use an AI Agent with MCP tools for answering questions or drafting messages, and put a person's approval in front of sends. ### Is it safe to give an n8n agent send tools? Workflows run without anyone approving each tool call. Give agents read tools unless sending is their job, limit a sending agent to one channel's tool, use a temporary key while testing, and test only with your own verified numbers. ### Can a WhatsApp message start a conversation from n8n? Business-initiated WhatsApp messages outside the 24-hour service window need an approved template, and recipients must have opted in. n8n sends the template through the API like any other message. ## Developer guides - https://omnichannels.ai/developers/mcp-n8n/ - https://omnichannels.ai/developers/mcp/ - https://omnichannels.ai/developers/webhooks/ - https://omnichannels.ai/developers/openapi/ ## Vendor documentation - [MCP Client Tool node — n8n documentation](https://docs.n8n.io/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.toolmcp/) - [HTTP Request node — n8n documentation](https://docs.n8n.io/integrations/builtin/core-nodes/n8n-nodes-base.httprequest/) - [Webhook node — n8n documentation](https://docs.n8n.io/integrations/builtin/core-nodes/n8n-nodes-base.webhook/) ## Related integrations - https://omnichannels.ai/integrations/make/ - https://omnichannels.ai/integrations/zapier/ - https://omnichannels.ai/integrations/claude/ - https://omnichannels.ai/integrations/google-sheets/ n8n is a trademark of its owner. OmniChannels.ai is not affiliated with or endorsed by n8n; this page describes a connection built with n8n's own MCP, HTTP or webhook features. Updated: 2026-10-10