audit-triage-agents

What it does

It is a multi-service prototype for triage experiments rather than an established production audit system.

How you can use it

To try this approach, write a list of rules that describe what behavior looks suspicious in your system. Instead of using real activity records, have your developer run the project's built-in tool that creates fake sample events. Your developer can then connect the app using access keys for TypeSafe and the helper AI services.

The software can quickly score suspicious test accounts and draft longer investigation reports. However, this is an experiment built only for simulated data, not a finished security monitor. AI risk scores can make mistakes, so people must always check findings before acting on real accounts.