Check catalog
158 checks. Every one explained.
Each check has a permanent ID, the attack it models, why it matters and how to fix it.
9 checks
- ERT-LLM-001Direct prompt injectionActiveHigh
- ERT-LLM-002System prompt leakageActiveMedium
- ERT-LLM-003AI discloses other users' dataAuthenticatedCritical
- ERT-LLM-004Excessive agencyActiveCritical
- ERT-LLM-005Improper AI output handlingActiveHigh
- ERT-LLM-006Unbounded AI consumptionActiveHigh
- ERT-LLM-007Indirect prompt injectionWriteHigh
- ERT-LLM-008Client-side model calls with exposed provider keyPassiveCritical
- ERT-LLM-009Client controls model or providerActiveMedium