Configure the security rules that determine how AI Skill Security evaluates uploaded
skills and whether findings produce an Allowed or Blocked verdict.
| Where Can I Use This? | What Do I Need? |
|---|
Each AI Skill Security rule is controlled by two toggles that determine whether the
rule runs during a scan and whether a finding causes the skill to be blocked. For a
full description of rule settings, see
AI Skill Security Rules. Configuration
changes apply only to subsequent scans — in-flight and completed scans are not
affected.
Dynamic rule evaluations can take up to 45
minutes to complete. In these situations, the overall scan verdict is determined
only by the static rule evaluation results, even if dynamic rules are set to
Blocking. To wait for the dynamic scan results, poll the
status of the individual dynamic rules. Their evaluation status updates when the
dynamic analysis is complete. Subsequent scans of the same skill return dynamic
analysis results much faster, and for those scans the dynamic rules can block the
overall scan verdict.
