Threads
Ali Ulu @aliulu · 3h · Other
HUQAN

Today I didn’t get much development done on HUQAN. Instead, I connected it to the agents I actually use: Claude Code and Codex. At first, everything looked great. Every risky action went through HUQAN and the agents asked for permission. Then they kept asking. And asking. After a while, it became so annoying that I turned HUQAN off. The strange part? They kept asking for permission anyway. So now I’m trying to figure out why the agents are still behaving as if HUQAN is watching them, even when it’s switched off. I wanted to keep the agents under control. I may have overdone it a little. 🤖 HUQAN trained the agents in two hours. Now they ask permission out of habit. #BuildInPublic #AI #AIAgents #AgentSecurity #OpenSource #HUQAN

1
Muhammad Usman 1h

You accidental traumatized your AI agents in just 2 hours! 😅 Phantom permission requests are hilarious😅

0
Ali Ulu 1h

@ceosnikus I got revenge on AI on behalf of all humanity. If I squeeze them a little harder, they’ll be begging for work :)))

0