The Agent Skills Directory

[Prompt Injection] (SAFE): The instructions are strictly limited to creative scriptwriting. There are no attempts to bypass safety filters or override system behavior.
[Data Exposure & Exfiltration] (SAFE): No access to sensitive files, environment variables, or hardcoded credentials was found. There are no network-bound operations.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill contains no code, shell commands, or external package requirements. It operates entirely within the LLM's text generation context.
[Indirect Prompt Injection] (LOW): User input is interpolated via $ARGUMENTS without explicit boundary markers. While a user could potentially provide input that influences the generated script in unintended ways, the skill lacks the capabilities (such as file system or network access) to turn such an injection into a security threat.

short-vid