The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied data, including photos and descriptive text, creating a potential surface for indirect prompt injection.
Ingestion points: The skill ingests a user photo via upload_asset and user-provided reaction descriptions in SKILL.md to guide the image-to-image and image-to-video processes.
Boundary markers: The skill does not implement explicit prompt delimiters or instructions to the agent to ignore potentially malicious instructions embedded in the input data.
Capability inventory: The skill utilizes various generative and assembly tools including scenario-image, scenario-video, scenario-video-assembly, and scenario-asset-analysis to perform file operations and generate content.
Sanitization: It employs a multi-stage validation process using scenario-asset-analysis to compare generated frames against the original reference photo and mentions scenario-moderation for filtering content.

scenario-fan-cam