The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions define an interface that ingests external data, creating a potential surface for indirect prompt injection attacks.
Ingestion points: Data enters the agent context through the prompt, image_url, reference_video_urls, audio_url, and media parameters described in SKILL.md.
Boundary markers: The skill does not define specific delimiters or instructions for the model to ignore potential commands embedded within the media or prompt text.
Capability inventory: The skill enables the agent to perform network requests to external API endpoints using curl and JSON payloads.
Sanitization: There are no explicit sanitization, validation, or filtering mechanisms described for the external content before it is processed.
[EXTERNAL_DOWNLOADS]: The skill refers to the mcp-wan package and makes network requests to api.acedata.cloud. These resources are provided by the skill author, acedatacloud, and represent official vendor infrastructure for the service.

wan-video