Tutorial: connecting n8n AI Agent to live news via MCP and HTTP

On November 13, 2026, DEV Community's AI channel published a tutorial on connecting n8n's AI Agent node to live news and web search, outlining two practical approaches. The guide points out that the AI Agent node is strong at reasoning over the material it is given, yet if it is asked about something that happened that morning it can only guess, because the model behind it stopped learning months ago. That gap is the reason a workflow needs a way to pull in current information on its own.

The two approaches target different scenarios. In Agent mode, an MCP server is connected to the AI Agent node so the agent itself decides when to search the news, search the web, or read an article in full; this fits chat-style workflows and research-heavy tasks best. In Pipeline mode, a scheduled workflow issues ordinary HTTP requests to gather the content. Both examples in the guide rely on the same NestDaddy search API, which offers a 7-day free key with 5,000 calls included, enough for users to set up and test a working flow.

As of the available reporting, the guide stops at describing these two integration routes and the test setup. No follow-up updates on the two methods, the NestDaddy key limits, or how the approaches perform in practice have been published yet.