The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted markdown files which could contain malicious instructions designed to influence the agent's behavior during the conversion pipeline.
Ingestion points: The skill reads content from local markdown files specified by the user via the --input argument in scripts/md2pdf.py.
Boundary markers: The instructions do not define delimiters or specific guidance to ensure the agent ignores natural language commands embedded within the input markdown.
Capability inventory: The skill possesses capabilities for file system operations (reading markdown and writing PDF output), network access (fetching remote images), and instructions guide the agent to execute shell commands for the Python script and Pandoc.
Sanitization: There is no evidence of sanitization or filtering to prevent the interpretation of instructions found within the input data.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch external assets required for PDF generation.
Fetches Twemoji images from Cloudflare's CDN (cdnjs.cloudflare.com), which is a well-known service.
Uses urllib.request in scripts/md2pdf.py to download remote images specified by URLs in the input markdown documents.
[COMMAND_EXECUTION]: The skill's instructions in SKILL.md guide the agent to execute shell commands, including running the conversion script and a Pandoc-based fallback engine for complex documents.

lov-any2pdf