Policy Center
Admin-only controls for risk rules, policy templates, thresholds, and policy simulation
Public demo mode: read-only. User Profile switching simulates access scope and is not production authentication.
Policy Templates
0
0 templates enabled for active policy bindings
Enabled Rules
0
0 unique rules across policy templates
24h Rule Hits
0
Total rule hits across applications and policies
High False-Positive Rules
0
Rules with human-reviewed false-positive rate above 15%
Policy Templates
Package rules, thresholds, and default actions by AI application type
No policy templates returned by the API.
Seed policy templates or check the Policy Center API before tuning rules.
Rule Library
Manage rule state, base score, default action, hit count, and human-reviewed false-positive rate
No policy rules returned by the API.
Rule Library no longer falls back to frontend sample rules.
Template Rule Tuning
Per template, set the score bands that apply to every call, then tune individual rules: the action a rule takes and the score from which that action applies.
Score Bands
Enforced on every call from applications bound to this template. Bands only raise the response that matched rules already call for.
This template has no rules.
Policy Simulation
Test a single AI call against the current rule engine before changing policy behavior.
Detection Sandbox
Validate rule matches, risk score, evidence chain, and response action with a single AI-call sample.