SkydiveAI coworkers that actually get work done
Promoted
Chloros — AI agent security
Security for your own AI agents — before they act, not after
4 followers
Security for your own AI agents — before they act, not after
4 followers
You're running AI agents that send email, move money, touch files. The moment an agent can act, a bad instruction becomes a bad action — with no undo. Most tools log what happened after the fact. Chloros is a gate outside your agent that checks each action before it fires: scan skills before you install, hold irreversible actions (money, sends, deletes) for your approval, rewind or kill when needed. Built for people running their own agents — not the enterprise. Free to try.
A dry-run mode would be huge, like a simulation view that shows exactly what an agent plans to do before any approval prompt fires. Helps catch weird side effects from chained skills without putting real actions on pause every single time.
@mertcana2lr Great suggestion — a dry-run/simulation view that shows the full planned action chain before anything fires, so you can catch weird side effects from chained skills without pausing every action. That's a real blind spot, and it's exactly the kind of thing I want to build. Adding it to the list — thank you, genuinely useful.
Curious: is it the multi-step chains that worry you most, or single actions with non-obvious downstream effects?