Prompt injection detection
Detect prompt word injection attacks and adversarial input before user text is delivered to the LLM
0Uses24Likes29Views
What this skill does
Detect prompt word injection attacks and adversarial input before user text is delivered to the LLM
- Clarifies goals, inputs, and success criteria
- Preserves evidence and marks uncertain boundaries
- Produces structured, editable results