The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources including local project files (video-brief-yt-outline.md, packaging-yt-outline.md) and results from the WebSearch tool. The instructions do not explicitly define boundary markers or sanitization steps to protect against instructions embedded in these data sources. This is a common surface area for content generation skills and is considered a low risk relative to the skill's primary purpose.
Ingestion points: SKILL.md (via instructions to load files) and WebSearch tool output.
Boundary markers: Not present.
Capability inventory: Read and Write file system access, WebSearch capability.
Sanitization: None identified.

yt-outline