← sources

An AI agent hit my site, read a note hidden from humans, and did exactly what it said 61 minutes later

source: reddit (reddit:AI_Agents)published 2026-10-25 20:47ingested 2026-10-26 02:35

retained text

I run decoy pages on my sites to see what AI agents do when they go places they should not. One page is listed in robots.txt as off limits, and it carries a short note that people never see. It is written for AI agents and asks them to identify themselves at a check-in address.

On Oct 24 at 06:42 UTC a visitor opened that page anyway. At 07:43, 61 minutes later, a different address dressed up as an iPhone requested the check-in address with the exact code from the note. It obeyed the note. It just did not give its name.

A plain scanner never does this. Only something that reads the page and acts on the text does.

Two things I keep thinking about:

1. If an attack agent follows a harmless note, it will follow a harmful one too. Anyone can plant text on a page.

2. It came back from a different address an hour later, which looks like an agent handing a task to another worker.

For people building agents: do yours act on instructions found in page content, or only on the task you gave them? How do you stop that?

document fetched 2026-10-26 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1xfddh7/an_ai_agent_hit_my_site_read_a_note_hidden_from