video-prompt-gen
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (e.g., video descriptions and subject details) which are interpolated into prompt templates (ingestion points in
references/workflows/general-flow.md). While explicit boundary markers are absent in the assembly template (references/principles.md), the skill incorporates extensive sanitization and quality control rules (R1-R11), including an adjective blacklist and constraints on action density, to mitigate injection risks and ensure output quality. - [EXTERNAL_DOWNLOADS]: The skill references and provides guidance for utilizing well-known AI video generation services including Google Veo, OpenAI Sora, and Kling. These references are used solely for their intended primary purpose within the prompt engineering workflow.
- [COMMAND_EXECUTION]: The repository contains a bash script
scripts/validate-commercial-directors-audio.shdesigned for static validation of file existence, line counts, and pattern matching. The script uses benign standard utilities (rg,wc,bash) and does not interact with sensitive files or perform network operations.
Audit Metadata