The Agent Skills Directory

[SAFE]: The skill instructions and reference materials focus entirely on providing structured prompt templates for video generation. No evidence of malicious intent, command execution, or credential theft was found.
[INDIRECT_PROMPT_INJECTION]: The skill acts as a prompt engineering assistant that ingests user-provided information to generate complex video prompts, representing a standard indirect injection surface.
Ingestion points: The workflow in SKILL.md collects user inputs for subjects, products, settings, and other video parameters.
Boundary markers: The skill lacks specific instructions for the agent to ignore or sanitize potentially malicious instructions embedded within the user's sports descriptions.
Capability inventory: The skill is restricted to text generation for video prompts and does not possess capabilities for file writing, system command execution, or network data exfiltration.
Sanitization: No input validation or escaping mechanisms are described in the instructions.
[EXTERNAL_DOWNLOADS]: The skill refers to external documentation and verified case studies hosted on the author's GitHub repositories and the GoodCase gallery.
References content from github.com/LearnPrompt/awesome-seedance and github.com/LearnPrompt/goodcase-lite for daily refreshed case data.
Links to goodcase.ai and x.com provide media evidence and original sources for the prompt examples used in the skill.

seedance-sports-extreme