The Agent Skills Directory

[SAFE]: The skill provides legitimate documentation and workflow guidance for using video generation models via a specific service provider (Scenario).
[EXTERNAL_DOWNLOADS]: The skill mentions installing sibling skills using npx skills add scenario-labs/skills. As this refers to the official repository of the skill author (scenario-labs), it is classified as a standard vendor resource for extending platform functionality.
[COMMAND_EXECUTION]: The skill describes using standard MCP tools such as model_run, search, and upload_asset to interact with a remote API. All command parameters described follow documented API schemas for media generation.
[INDIRECT_PROMPT_INJECTION]: The skill identifies that user-provided text (prompts) is interpolated into model parameters. While this is an attack surface, the instructions focus on technical parameter formatting and specific model requirements rather than bypassing safety controls.

scenario-wan