baoyu-slide-deck

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local TypeScript scripts (scripts/merge-to-pptx.ts and scripts/merge-to-pdf.ts) using the bun runtime via npx. These scripts are used to aggregate generated slide images into final PPTX and PDF documents. The operations are limited to reading images from a specific project directory and writing the output files.
  • [EXTERNAL_DOWNLOADS]: The document generation scripts utilize standard libraries (pdf-lib and pptxgenjs) which are typically fetched from the public NPM registry. These are legitimate, widely-used packages for their respective purposes.
  • [PROMPT_INJECTION]: The skill processes untrusted user-provided content to generate slide outlines and image generation prompts, which constitutes a surface for indirect prompt injection.
  • Ingestion points: User-provided text or markdown files entered at the start of the workflow.
  • Boundary markers: The skill uses a clear templating system (references/base-prompt.md) with distinct sections for style instructions and slide content to provide structure to the model prompts.
  • Capability inventory: The skill facilitates image generation through an external capability and performs file write operations to the local filesystem.
  • Sanitization: While no automated sanitization of the user input is described, the skill enforces a multi-round manual review process (Steps 2, 4, and 6) where the user must explicitly approve the generated outline and prompts before image generation occurs, providing a strong human-in-the-loop defense.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 03:19 AM
Security Audit — agent-trust-hub — baoyu-slide-deck