The Agent Skills Directory

[SAFE]: The skill instructions define a legitimate process for video composition, including media ingestion, editing steps, and final delivery using platform-specific tools.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data in the form of screen recordings and text descriptions to generate video content. (1) Ingestion points: import_media for recordings and user-provided text inputs mentioned in SKILL.md. (2) Boundary markers: No specific delimiters or "ignore" instructions are defined for the processed user content within the skill. (3) Capability inventory: The skill utilizes tools like import_media, search_resources, and apply_composition to manipulate and export media. (4) Sanitization: No explicit sanitization or validation of the input media's metadata or transcripts is mentioned.

launch-video