portal
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a local Node.js script using the command
node ~/.claude/skills/video-effects/run.cjs. This is the primary functional requirement for applying the video effect.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a data ingestion surface by accepting and processing user-provided MP4 video files.\n - Ingestion points: User-provided MP4 video clips (
my-clip.mp4) specified inSKILL.md.\n - Boundary markers: None present in the skill instructions.\n
- Capability inventory: Execution of a shell command via
nodeto process the file.\n - Sanitization: No specific sanitization or validation of the video file metadata is described.
Audit Metadata