nature-paper2ppt
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from scientific papers, which presents an inherent surface area for indirect prompt injection. However, its capabilities are scoped to content summarization and document creation.
- Ingestion points: Processes PDFs, preprints, abstracts, and figure legends from various scientific sources (SKILL.md, README.md).
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified for the input data processing.
- Capability inventory: Uses PyMuPDF for reading, Pillow for image manipulation, and python-pptx for file writing (SKILL.md).
- Sanitization: No specific sanitization or validation of the input text content is mentioned.
- [COMMAND_EXECUTION]: The instructions guide the agent to perform environment setup, including creating virtual environments and executing installation commands for the required toolchain (SKILL.md).
- [EXTERNAL_DOWNLOADS]: The skill relies on external Python packages (PyMuPDF, Pillow, python-pptx). The workflow includes potentially downloading these from standard registries if they are not present in the execution environment.
Audit Metadata