embedded-captions
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: Analysis of the instructions and markdown documentation reveals no attempts to bypass safety filters or override agent constraints.
- [DATA_EXFILTRATION]: No access to sensitive user files or suspicious network activity was identified. Network requests are limited to fetching fonts and scripts from well-known sources.
- [REMOTE_CODE_EXECUTION]: The skill uses local subprocesses to run trusted utilities like ffmpeg and node. No unsafe evaluation of dynamic strings or execution of remote scripts was found.
- [EXTERNAL_DOWNLOADS]: Fetches library scripts and font assets from reputable sources such as jsdelivr and utilizes standard package managers for its primary functions.
Audit Metadata