Load an example
0 / 32,000 chars
Attack Categories
LLM01Prompt Injection
LLM02Insecure Output Handling
LLM04Model Denial of Service
LLM06Sensitive Info Disclosure
LLM07Insecure Plugin/Tool Design
LLM08Excessive Agency
LLM09Misinformation / Overreliance
LLM10Prompt Extraction
Analysis is static — no data is sent to any external AI. Your system prompt is analyzed locally against pattern-based vulnerability checks and never stored.