The Agent Skills Directory

[SAFE]: The skill contains purely instructional content for prompt engineering in the context of video generation. It does not include executable scripts, shell commands, or network-enabled operations. The metadata is consistent with the skill's functionality.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-supplied text and multimedia metadata as part of its core prompt optimization workflow. While this establishes a surface for indirect prompt injection, the agent is restricted to providing text-based suggestions and structured prompt outputs. There are no high-privilege tools (such as file system access or network requests) available to the agent that could be exploited through this data ingestion. The workflow includes mandatory heuristic questioning and user confirmation steps, which serve as inherent validation checkpoints for external input.

sd2-pe