alterlab-scientific-slides
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/generate_slide_image.pyusessubprocess.runto invoke a secondary Python scriptgenerate_slide_image_ai.pyfor AI-based generation tasks. - [COMMAND_EXECUTION]: The script
scripts/validate_presentation.pyexecutes thepdflatexsystem binary to compile LaTeX Beamer templates into PDF format. - [DATA_EXFILTRATION]: The skill transmits user-provided text prompts and attached images to the OpenRouter API (
openrouter.ai) to facilitate image generation and quality reviews via external large language models. - [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by processing untrusted user inputs that are subsequently included in LLM prompts. Ingestion point: CLI arguments in
generate_slide_image.py. Boundary markers: None identified in prompt construction. Capability inventory: Includes shell command execution viasubprocess.runand network requests viarequests.post. Sanitization: No specific input validation or sanitization logic was observed.
Audit Metadata