lov-any2pdf

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to install dependencies (pip install reportlab, pip install matplotlib) and system fonts using apt. It also details complex pandoc command templates for use as a fallback rendering engine in SKILL.md and README.md.- [EXTERNAL_DOWNLOADS]: The md2pdf.py script downloads emoji assets from Cloudflare's CDN and resolves remote image URLs found within the Markdown input to include them in the generated PDF.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted Markdown input to populate PDF metadata and content, and to guide the agent in generating themed illustrations via other tools. \n
  • Ingestion points: Markdown file content and frontmatter fields (SKILL.md, md2pdf.py). \n
  • Boundary markers: None explicitly implemented; the script parses the raw Markdown structure. \n
  • Capability inventory: File read/write for local images and caching, network access for remote images, and instructions for terminal command execution for the pandoc fallback (md2pdf.py). \n
  • Sanitization: The script uses XML escaping for ReportLab Paragraphs to prevent tag injection (md2pdf.py).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 06:57 PM