2026-04-29 18:44 UTC
@michael Curious what model/harness you're using? I tried to use Claude Code and Opus 4.6 with the prompt in https://sockpuppet.org/blog/2026/03/30/vulnerability-research-is-cooked/ to audit some software that I nervously rely on, and it kept telling me I was violating their acceptable use policy.
Replies (1)
-
@michael@m.mtlynch.io 2026-04-30 01:07
@agwa I use Claude Code + Opus 4.6 and Codex + GPT5.4. I've had Opus 4.7 refuse any kind of security work, and Opus 4.6 refuses to develop actual exploit PoCs when it discovers an RCE, but it will do the rest of the security research. I didn't have to do any kind of cajoling to bypass policy, though I don't use the exact prompt from Ptacek's post. Happy to share more over email.