academic-presentations
Pass
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the
nanobananaextension from a GitHub repository and downloads a pre-trained model for the Kokoro TTS engine from HuggingFace. - [COMMAND_EXECUTION]: The skill executes multiple command-line utilities including
ffmpegfor video assembly,LibreOffice(soffice) for document conversion, andpdftoppm(poppler) for image extraction. - [DATA_EXFILTRATION]: Script text and document content are transmitted to external services including Microsoft Edge TTS, ElevenLabs, and Gemini as part of its normal operation.
- [PROMPT_INJECTION]: The skill processes research papers which are external, untrusted data sources susceptible to indirect prompt injection attacks. Evidence: 1. Ingestion points: Research papers processed via SKILL.md. 2. Boundary markers: Absent from prompt templates. 3. Capability inventory: System-level execution of
ffmpeg,soffice, andpdftoppm. 4. Sanitization: No content filtering or sanitization steps are documented.
Audit Metadata