yt-script
Pass
Audited by Gen Agent Trust Hub on Nov 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input in the form of a "raw idea" to generate YouTube scripts and hooks. This content is subsequently passed to a local Python script for scoring, creating an injection surface.
- Ingestion points: The user's "raw idea" described in
SKILL.mdis used as the basis for all generated content. - Boundary markers: The instructions do not define clear delimiters or provide the agent with guidance to disregard instructions embedded within the user-provided idea.
- Capability inventory: The skill executes a local script
hookscore.pyand reads from both the current project and a skill-specific directory in the user's home folder (~/.youtube-agent-skill/voice.md). - Sanitization: There are no explicit sanitization, filtering, or validation steps applied to the user input before it is interpolated into prompts or processed by the scoring tool.
Audit Metadata