pdf

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill includes plaintext password literals in example commands/code (e.g., qpdf --password=mypassword and writer.encrypt("userpassword","ownerpassword")), which requires the LLM to reproduce secret values verbatim in outputs.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow reads outsider-authored free text embedded in user-provided PDFs by extracting labels/text/annotations (e.g., extract_form_structure.py uses pdfplumber page.extract_words() and fill_pdf_form_with_annotations.py reads entry_text["text"] from user-supplied fields.json to create PDF FreeText annotations).