nature-paper2ppt

Pass

Audited by Gen Agent Trust Hub on May 25, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by processing untrusted scientific papers (SKILL.md). Ingestion points: Accepts paper PDFs, preprints, and article text as inputs. Boundary markers: Absent; there are no specific markers used to separate untrusted source content from instructions. Capability inventory: The skill utilizes python-pptx to generate local files and may execute shell commands for rendering (LibreOffice) or environment setup. Sanitization: Absent; no sanitization or escaping of extracted paper content is mentioned before it is used to populate slide text.
  • [REMOTE_CODE_EXECUTION]: The skill specifies the installation of well-known Python packages (PyMuPDF, Pillow, python-pptx) when they are missing from the runtime environment. These are standard libraries for document processing and are documented neutrally.
Audit Metadata
Risk Level
SAFE
Analyzed
May 25, 2026, 08:08 AM
Security Audit — agent-trust-hub — nature-paper2ppt