The Agent Skills Directory

[SAFE]: The skill utilizes well-established Python packages (pdfplumber, weasyprint, and reportlab) to perform document translation and typesetting. The implementation is transparent and follows standard practices for document processing tools.\n- [COMMAND_EXECUTION]: The skill includes scripts for local execution and configures environment variables (such as DYLD_FALLBACK_LIBRARY_PATH) required for cross-platform library support. These operations are necessary for the document generation engine and do not represent a security risk.\n- [PROMPT_INJECTION]: The skill processes untrusted external content by extracting text from PDFs using pdfplumber, which creates a surface for indirect prompt injection. Ingestion points occur in scripts/translate_pdf.py and SKILL.md (Step 1). Capability inventory includes file-system write operations through weasyprint and reportlab. While sanitization and boundary markers are absent, the skill itself contains no malicious override instructions or attempts to bypass safety filters.

pdf-translate