Launching soon — Windows · macOS next
Your AI agents have full access. It's time to control them.
See what each agent can reach on your machine—then fix weak configs fast.
Windows first · macOS next
See what each agent can reach on your machine—then fix weak configs fast.
Windows first · macOS next
In the desktop app
Free: detection, scores, guided fixes, freeze on configs, audit trail. Premium: auto-fix, Fix All, Guard AI.
Finds supported local AI agents, reads their configs, and shows what they’re allowed to do.
Severity rating + plain-English explanation for every finding.
Automatically apply safer values to config files where Cuper supports it — per finding or bulk across agents. Free tier uses step-by-step guided fixes instead.
One button backs up and rewrites detected agent configs to turn off the riskiest flags. Reversible. Does not kill running apps.
Now: GPT-powered answers based on your scan and findings, in plain language.
Next: Security that understands your live environment and can act.
Every scan, fix, and decision — timestamped and logged. Full accountability.
The Roadmap
Start, stop, and guide agents from your phone.
Tell agents exactly what to do. Real-time execution.
Runtime enforcement: approve or deny actions as they happen — beyond today’s config scanning.
Every AI system in one dashboard. Zero chaos.
Download. Run a scan. Every AI agent found in seconds.
Shell access, file permissions, network — each risk explained clearly.
One-click fix where supported. Optional emergency freeze on configs. Re-scan to confirm.
No card required
or $120/yr — save 17%