hyperframes-embedded-captions
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes transcription data generated from user-provided video/audio files. This transcript is interpreted by the AI agent to make layout and typographic choices. The skill includes explicit instructions for the agent to validate transcripts before use, and the rendering logic applies standard HTML escaping to text tokens before interpolation into the DOM.
- [EXTERNAL_DOWNLOADS]: The skill references the GSAP animation library from the jsDelivr CDN and utilizes
uvxto execute thewhisperxtranscription tool. It also suggests updates vianpx. These external dependencies originate from established tech companies and public registries. - [DYNAMIC_EXECUTION]: The skill generates local shell scripts (
_postfx.sh) to apply video effects such as noise, shake, and grain. These scripts are created from deterministic templates based on parameters defined in the skill's theme files. - [COMMAND_EXECUTION]: The skill orchestrates multiple local CLI utilities, including
ffmpeg,ffprobe, andnode, for media processing tasks. These operations are conducted through structured process spawning and are consistent with the skill's stated purpose of local video manipulation.
Audit Metadata