storyboard-shot-breakdown
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external media and transcripts, creating a surface for indirect prompt injection. 1. Ingestion points: Workflow steps 1 and 2 in SKILL.md involve reading project state, source media, and transcript context. 2. Boundary markers: The skill does not define specific delimiters to isolate external content. 3. Capability inventory: The skill calls media tools like OpenChatCut and may execute external scripts. 4. Sanitization: No sanitization of ingested media metadata or transcript text is specified.
- [COMMAND_EXECUTION]: The workflow involves invoking external media utilities for shot boundary detection and frame extraction. The skill includes a specific rule to prefer established tools over hardcoded local commands, which is a positive security practice for maintaining environment stability.
Audit Metadata