md-to-pptx

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill parses and processes user-provided Markdown files. If these files are obtained from untrusted sources (e.g., downloaded from the web or provided in a PR), they could contain natural language instructions intended to influence the agent's behavior when it reads the file to generate outlines or execute the conversion.
  • Ingestion points: The content of the Markdown file (input.md) is ingested by the agent and parsed by scripts/md_parser.py.
  • Boundary markers: The skill does not implement explicit delimiters or instructions to the agent to treat the Markdown content as untrusted data.
  • Capability inventory: The skill can execute local scripts and has file system access to read images and write the resulting PPTX files.
  • Sanitization: While the parser removes Markdown syntax for formatting, it does not sanitize the semantic content for instruction injection patterns.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of python-pptx and Pillow via pip. These are well-known, legitimate libraries for PowerPoint generation and image processing, representing standard dependency management for this use case.
  • [COMMAND_EXECUTION]: The conversion process involves running a local Python script (scripts/md_to_pptx.py). This is the primary function of the skill and is triggered by the agent within its execution environment to process local documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 11:05 AM