pdf

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes examples that embed passwords directly in commands and code (e.g., qpdf --password=mypassword and writer.encrypt("userpassword","ownerpassword")), which requires generating or handling secret values verbatim and is a high exfiltration risk.

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime workflow for non-fillable PDFs ingests outsider-authored free text from fields.json (specifically entry_text.text) into the LLM context via the agent’s user-provided JSON/field values, and that text is then rendered into PDF FreeText annotations.