The Agent Skills Directory

[SAFE]: The skill provides structured guidance and documentation for a multi-step generative AI pipeline, mapping user requirements to specific tool sequences.
[COMMAND_EXECUTION]: Includes instructions to use npx for installing related components from the author's own repository (PrunaAI/pruna-skills). These are standard package management operations for the platform.
[COMMAND_EXECUTION]: Mentions the execution of local Python scripts (e.g., run_from_plan.py) to manage generation workflows and plan runners, which is consistent with the intended use of the pipeline.
[EXTERNAL_DOWNLOADS]: References external models and tools (such as Replicate) that are required for specific audio and image tasks. These references include standard guidance on API token requirements.
[SAFE]: Implements a "Staged generation gate" system as a security and cost-management best practice, requiring human-in-the-loop approval before executing high-cost video generation or final assembly tasks.

recipe-catalog