The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses ffmpeg and ffprobe via the Python subprocess module to extract identity frames from videos and to verify that generated images meet specific ratio and resolution requirements. These commands are invoked using list-based arguments rather than shell strings, which is a safe practice that prevents command injection.
[PROMPT_INJECTION]: The skill processes user-supplied data (such as titles, tags, and highlights) to build prompts for an image generation tool. While this represents an indirect prompt injection surface, the risk is negligible as the input is delimited with quotes and the capability is restricted to visual asset generation.
[SAFE]: The skill follows a modular and transparent design. It stores style configurations and gesture presets in structured JSON manifests and uses local assets for generation. No obfuscation, hardcoded credentials, or external network requests were detected during the analysis.

skill-cover