The Agent Skills Directory

[SAFE]: The skill documents a workflow for media creation using a suite of integrated AI tools. The instructions are consistent with the stated purpose of generating podcast videos.
[DATA_EXPOSURE]: No hardcoded credentials or attempts to access sensitive system files (e.g., SSH keys, AWS configs) were detected. Data handling is limited to processing media URLs provided by the user and saving temporary JSON files for tool coordination.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to interact with the infsh platform's CLI. These commands are used to invoke specialized AI applications for image training, generation, and media merging in a controlled manner.
[PROMPT_INJECTION]: There are no indicators of prompt injection, such as instructions to ignore safety filters or reveal system prompts. The 'Rules' provided are technical best practices for optimizing video output quality and preventing GPU memory errors.
[EXTERNAL_DOWNLOADS]: No suspicious external downloads or remote script executions (e.g., curl | bash) were identified. External resource usage is confined to official platform apps and well-known service providers.

ai-podcast