Hi PH, I'm Shashwat, building access control for AI agents

Hey everyone! I'm Shashwat from India, an information security student and founder of StartiQOS AI. AI agents are getting real access to databases, payment APIs and inboxes, and most teams still hand them a raw API key. I'm building a runtime gateway that grants agents access by policy and logs every action they take. I'll be building in public here over the next few weeks before we launch. Question for anyone running agents in production: what's the one tool or permission your agent has that makes you most nervous?