lov-any2pdf
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches emoji icon assets from Cloudflare's CDN (cdnjs.cloudflare.com) at runtime to handle emoji rendering in the generated PDF. This is a reference to a well-known service for legitimate assets.
- [INDIRECT_PROMPT_INJECTION]: The skill processes markdown files that may be provided by the user or external sources and automatically fetches images from remote URLs contained within the document. This represents a potential vulnerability surface:
- Ingestion points: The
scripts/md2pdf.pyscript ingests untrusted markdown data through the--inputargument, which is then parsed to identify images and other assets. - Boundary markers: There are no boundary markers or instructions to the agent to treat embedded URLs as untrusted; the parser automatically attempts to resolve and download them.
- Capability inventory: The script uses
urllib.request.urlopento perform network requests for both emojis (scripts/md2pdf.py:431) and markdown images (scripts/md2pdf.py:868). It also has the capability to write to the system's temporary directory for caching. - Sanitization: The
_resolve_image_sourcefunction does not restrict requests to specific domains or validate the content beyond checking file extensions. It also supports thefile://URI scheme, which could allow local file inclusion if a malicious path is provided in the markdown source.
Audit Metadata