The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could serve as a vector for indirect prompt injection.
Ingestion points: Data enters the agent context through the search_youtube and tavily-search tool calls described in SKILL.md.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded prompts within the retrieved search results or video metadata.
Capability inventory: The skill uses analysis and search tools; it does not utilize high-risk capabilities such as file writing, arbitrary command execution, or network exfiltration tools.
Sanitization: There is no evidence of sanitization, filtering, or validation logic for the external content retrieved by the search tools.

movie-analysis