using-document-designer

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill functions as a professional design assistant, providing educational content, typesetting templates, and configuration guidance for tools like Typst and Pandoc. All included Typst code snippets represent standard usage for document formatting.
  • [EXTERNAL_DOWNLOADS]: The skill references established community resources such as the Typst Universe package registry and the Citation Style Language (CSL) repository on GitHub. These are well-known services for the intended typesetting workflows.
  • Evidence: Mentions of typst.app/universe for templates and github.com/citation-style-language/styles for citation management.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text and external documents, which is a potential surface for indirect prompt injection. Given that the primary purpose is visual formatting and that the skill relies on standard platform guardrails, the risk is categorized as safe.
  • Ingestion points: The skill ingests user content in various formats (Markdown, DOCX, etc.) as described in SKILL.md.
  • Boundary markers: While not explicitly defined in the provided templates, the Typst and Pandoc syntax structures content separately from layout logic.
  • Capability inventory: The agent utilizes the typst and pandoc command-line tools for document conversion and PDF generation.
  • Sanitization: The skill relies on the underlying agent platform's safety filters to manage untrusted input data.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 04:26 PM