The Agent Skills Directory

[SAFE]: The skill documents a multi-step process for media creation using an integrated CLI tool (infsh). The workflow consists of character generation, text-to-speech synthesis, avatar animation, and video merging.
[COMMAND_EXECUTION]: The skill uses infsh app run to invoke specialized AI models (e.g., phota/train, inworld/text-to-speech-2, pruna/p-video-avatar). These operations are within the expected scope of a podcast generation skill and use structured JSON inputs for parameter passing.
[EXTERNAL_DOWNLOADS]: The phota/train tool accepts an array of image URLs provided by the user to train a facial identity profile. This is a standard functional requirement for consistent character generation and does not involve the execution of remote code or scripts.
[SAFE]: The instructions include specific quality assurance steps and performance guidelines (e.g., avoiding parallel GPU runs to prevent memory errors) which demonstrate a focus on operational stability rather than malicious intent.

ai-podcast