document-processing

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill instructs the agent to ask for a PDF password and shows using it as a literal string in code (reader.decrypt("the_password")), which would require the LLM to accept and embed a secret value verbatim in generated output/commands.

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow performs PDF/DOCX/XLSX/PPTX text extraction (e.g., pdfplumber/pypdf/OCR) from the user-provided document files, which can contain outsider-authored free text that is then ingested into the agent’s LLM context for downstream analysis.