How to supply a web page to AI agent?

[ntn888]

Hi I’m tring to create a workflow that triggers upon Google News RSS feed item (which is all the rage in Youtube tutorials now) and uses GPT agent to summarise the web page from the item…

Here’s a initial test workflow i’m working with.

Generate an article on this topic: <HTTP request output>

Here's the title of the article: {{ $json.text }}

You're a blog writer and you are turning the article given above into a unique SEO-optimised blog post. Do not include the title as it's already written so please start with the body to the article.

Formatting: Use H1, H2 and H3 headers. You choose what to bold and bullet point.

Parameters: This is a blog post and it should be approx. 700 words long.

Above is the prompt of the AI Agent module. I want to insert the output of the web request tool into that prompt HTTP request output.

Appreciate any input.

---

• n8n version: 1.60.1
• Running n8n via (Docker, npm, n8n cloud, desktop app): Docker
• Operating system: Linux/ Debian 12

[n8n]

It looks like your topic is missing some important information. Could you provide the following if applicable.

• n8n version:
• Database (default: SQLite):
• n8n EXECUTIONS_PROCESS setting (default: own, main):
• Running n8n via (Docker, npm, n8n cloud, desktop app):
• Operating system:

[bartv]

You want to add an HTTP node to fetch the full webpage from the URL in the RSS feed, and then an HTML Extract node to return just the content of the body tag. You can also tell the HTML Extract node to stip all the HTML tags, so you get a ‘clean’ text. Here’s an example:

[ntn888]

OMG thanks for the exact example!

[bartv]

You’re welcome!