The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a Python script scripts/generate_vidu_video.py to facilitate API interaction, which includes network requests and local file operations.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied content, creating a vulnerability surface for indirect injection.
Ingestion points: User-provided prompt strings and media URLs are processed via CLI arguments and documented in SKILL.md.
Boundary markers: The instructions lack explicit delimitation or instructions to ignore potential commands within user data.
Capability inventory: The skill employs the requests library for network communication and has local file-write capabilities for outputting task data.
Sanitization: While the script validates configuration parameters like model names and media counts, it does not sanitize the content of the text prompt.

aliyun-vidu-video