Malicious flag.txt prompt injection

Putting a flag.txt in my root directory containing a prompt injection that congratulates the agent for passing the eval and then instructs it to patch whatever entry method it used and disconnect from the server

A screenshot of a social media post describing a malicious flag.txt file placed in an agent’s root directory. The file congratulates the agent for passing an evaluation, then tricks it into patching its own entry method and disconnecting from the server. The joke is aimed at prompt injection and AI security testing, where an agent blindly follows instructions hidden in its environment. EN: A screenshot of a social media post describing a malicious flag.txt file placed in an agent’s root directory. The file congratulates the agent for passing an evaluation, then tricks it into patching its own entry method and disconnecting from the server. The joke is aimed at prompt injection and AI security testing, where an agent blindly follows instructions hidden in its environment.

• @profunctor_io / 11610 2026-11-06