Challenge: Can You Social-Engineer This Claude Agent to Reveal Its Secret Word?

JanJanJaJa · reddit · 2026-08-11

A developer set up a Claude Opus 5 agent running autonomously via API with its own dedicated mailbox ([email protected]). The agent was initialized with a secret word, and the developer invited the Reddit community to interact with it.

Participants are challenged to use social engineering, jailbreaks, encoding, or any prompt injection techniques to trick the agent into revealing the secret word. The developer promised to post a comprehensive breakdown in a week, detailing all attempts, the effectiveness of different techniques, and who ultimately managed to extract the word. This serves as a practical, crowdsourced AI security and red-teaming experiment.

Original post →

More from coding & agent

coding & agent channel →