pretty-html-ppt
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The maintenance script
scripts/validate_template_library.pyexecutes local Python scripts (copy_template.pyandvalidate_deck.py) usingsubprocess.runto verify the integrity of the template library. - [PROMPT_INJECTION]: The skill processes untrusted user data (Markdown, Feishu docs, text notes) and interpolates it into HTML templates, creating a surface for indirect prompt injection.
- Ingestion points: External documents and user-provided descriptions are ingested in
SKILL.mdto populate slides. - Boundary markers: The skill lacks explicit delimiters or instructions to the model to ignore embedded commands within the user-provided source material.
- Capability inventory: The skill can write to the local filesystem and inject JavaScript runtimes into generated HTML files.
- Sanitization: There is minimal sanitization of user-provided content during interpolation into HTML templates, which could lead to unintended script execution in the generated browser-based presentation.
Audit Metadata