render-cgi-sizzle
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a documentation and configuration 'recipe' for orchestrating a specific video production workflow. It defines parameters for asset compositing, audio mixing, and video finalization using industry-standard tools like FFmpeg and PIL (Python Imaging Library).
- [NO_CODE]: The documentation references several external scripts (e.g.,
render_eryn_vo.sh,gen_plates.py,composite_screens.py) required for the pipeline, but the skill package itself does not include the source code for these scripts. It describes the orchestration logic without providing the underlying executables. - [EXTERNAL_DOWNLOADS]: The pipeline description mentions legitimate integrations with third-party creative AI platforms, including ElevenLabs for voiceover/music and Fal.ai (nano-banana and Kling 3.0) for image and video generation. These are presented as separate capabilities orchestrated by the provided configuration.
- [PROMPT_INJECTION]: The instructions utilize strong imperative language ('ALWAYS real UI', 'NEVER AI-rendered') to enforce brand consistency and authenticity. These represent domain-specific constraints for the creative output and do not target the agent's safety filters or attempt to override its core operational guidelines.
Audit Metadata