baoyu-slide-deck

Pass

Audited by Gen Agent Trust Hub on Nov 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text from source files to generate slide outlines and image prompts. While this presents an attack surface for indirect prompt injection, the risk is mitigated by a mandatory confirmation policy and optional manual review steps for both the outline (Step 4) and the prompts (Step 6) before the image generation backend is invoked.
  • [COMMAND_EXECUTION]: The skill executes local TypeScript scripts (merge-to-pptx.ts and merge-to-pdf.ts) via the bun runtime to package generated images into final presentation formats. These scripts are part of the skill's core functionality for local file processing.
  • [DYNAMIC_EXECUTION]: The skill dynamically locates and executes a local wrapper script (codex-imagegen) if a native image generation tool is not found. It searches environment variables and the local plugin directory for this wrapper, which is a common pattern for managing vendor-specific toolchains in variable execution environments.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 2, 2026, 09:45 AM