The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of the @runcomfy/cli tool to interact with the RunComfy Model API. Documentation provides specific command examples for video generation tasks.
[EXTERNAL_DOWNLOADS]: The RunComfy CLI is documented to download generated media files from *.runcomfy.net and *.runcomfy.com. The skill notes a 2 GiB download size limit as a security measure.
[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it ingests user-provided text prompts and external media URLs (images, video, audio) for remote processing. 1. Ingestion points: Text prompt and media arrays (image_url, video_url, audio_url) passed to the model endpoint. 2. Boundary markers: Prompts are passed as JSON strings to the CLI to prevent shell injection (documented in 'Security & Privacy'). 3. Capability inventory: Executes CLI commands and performs network downloads to a user-specified output directory. 4. Sanitization: CLI uses HTTPS for transmission, JSON encoding for parameters, and enforces domain whitelisting and file size caps for output downloads.
[SAFE]: The skill documents secure credential handling practices, such as setting file permissions to 0600 for local token storage and supporting environment variables for CI/CD environments.

seedance-v2