Give your team a chat assistant that reads a web page and answers with clear steps. It is ideal for support, research, and product teams that need quick facts from live pages without switching tabs.

A chat message starts the flow. The agent uses OpenAI gpt 4o mini with a short memory window to keep context across turns. It calls a web scraping tool that fetches readable page text from the URL inside the message using r.jina.ai. The agent then analyzes the content and replies in simple language. No key is needed for the scraper, which makes testing easy. This design brings real time page data into the chat, so answers reflect the latest content on the site.

You need an OpenAI API key and prompts that include a URL. Expect faster answers from docs and better first reply quality. Use it to guide users with steps from product manuals, summarize release notes, or pull how to steps from knowledge articles. Activate the workflow and send a question with a full URL to see it work.

• Chat message trigger starts the conversation as soon as a user sends a prompt.
• Agent decides when to scrape and how to use the page content to answer.
• HTTP scraper pulls readable text from https://r.jina.ai/{url} with no API key.
• OpenAI gpt 4o mini generates clear and concise replies.
• Window buffer memory keeps recent messages for context across turns.
• Agent prompt guides URL extraction from the user message for the scraper.

• Reduce page reading time from 20 minutes to 2 minutes per question
• Automate up to 80 percent of link based inquiries in chat
• Improve answer accuracy by 30 percent by using live page content
• Handle 5 times more chat requests with the same team size
• Unify chat input and live web pages in one flow

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with OpenAI and Jina AI. See the Tools Required section above for links to create accounts with these services.
3. In the n8n credentials manager, create an OpenAI credential. Get your API key from the OpenAI account API page and paste it into the key field. Save the credential with a clear name like OpenAI Main.
4. Open the gpt 4o mini node and select your OpenAI credential in the Credential to connect with dropdown. Keep the model set to gpt 4o mini.
5. Open the Jina AI Web Scraper Tool node and confirm the URL is set to https://r.jina.ai/{url}. Verify the placeholder named url exists in Placeholder Definitions.
6. Open the Agent node and review the system prompt. Make sure it tells the agent to extract a URL from the user message and call the scraping tool when needed.
7. Open the Window Buffer Memory node and adjust the window size if you want more or fewer recent messages kept. The default works for most use cases.
8. Open the When chat message received trigger and ensure it is connected to the agent. Activate the workflow in the top right.
9. Test the flow by sending a message that includes a full URL and a question, for example Ask for steps from https://example.com/docs/page.
10. Check the execution log. Confirm the scraper ran and returned text, then verify the agent reply includes facts from the page.
11. Troubleshoot common issues: if no content is returned, make sure the URL starts with https and the site is publicly accessible. If you see OpenAI errors, verify your API key, quota, and model selection.
12. For best results, point to pages with clear text content. Avoid pages that require login or heavy client side rendering.