cutagent-real-estate-reel
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data such as property footage and supplied evidence which are potential vectors for indirect prompt injection.
- Ingestion points:
SKILL.mdspecifies reading "supplied evidence" and "supplied property information" to verify claims and guide the editing process. - Boundary markers: The instructions do not define explicit delimiters (e.g., XML tags or block markers) to separate these external inputs from the agent's instructions.
- Capability inventory: The skill references "SDK declarations" and "CutAgent CLI command cards," implying the use of command-line tools for video processing, rendering, and file operations.
- Sanitization: No explicit sanitization or instruction-ignoring patterns are defined for the ingested property data.
Audit Metadata