The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the renku CLI tool for various media generation lifecycle tasks, including models list, models show, schema show, and execute. These commands are integral to the skill's function and align with the author's toolset.
[EXTERNAL_DOWNLOADS]: Technical documentation and API schemas are referenced from fal.ai. These links provide necessary context for valid request authoring and do not involve executable code downloads.
[INDIRECT_PROMPT_INJECTION]: The skill processes external model guides and API schemas to determine request parameters. It mitigates injection risks by enforcing a specific JSON envelope structure for file markers and requiring manual confirmation steps within the Media Producer framework.
[SAFE]: The instructions incorporate multiple security best practices, including explicit prohibitions on hardcoding credentials, direct API access, and unauthorized output persistence.

fal-ai-media-provider