batch-convert

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill interacts with the host system by executing external command-line utilities including Pandoc, Marp, and LibreOffice via the subprocess module. Although the implementation uses list-based arguments to mitigate shell injection, the tool remains susceptible to argument/flag injection if input filenames or paths are not properly validated.\n- [INDIRECT_PROMPT_INJECTION]: The skill acts as a data ingestion pipeline, processing various document formats that may contain instructions targeting the AI agent.\n
  • Ingestion points: The batch_convert and convert methods in SKILL.md process external files.\n
  • Boundary markers: None identified; instructions within documents are not explicitly delimited or ignored.\n
  • Capability inventory: File system write access and external command execution.\n
  • Sanitization: No validation or sanitization of document content is implemented.\n- [EXTERNAL_DOWNLOADS]: The installation guide recommends fetching dependencies from standard package registries and official repositories, such as the markitdown library from Microsoft, pdf2docx from PyPI, and system-level tools like pandoc and libreoffice via official package managers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 06:57 AM
Security Audit — agent-trust-hub — batch-convert