lov-any2docx

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s required runtime path reads the user-provided Markdown file content at execution time (scripts/md2docx.py: main() reads args.input into md_text, then DocxBuilder.build() parses that md_text and embeds parsed text into the DOCX, with inline markdown/link text processed by the LLM none—but the workflow still ingests arbitrary free text authored by the requester).

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (medium risk: 0.40). The skill recommends running "pip install python-docx --break-system-packages" which explicitly suggests bypassing package-system protections and could modify system-installed Python packages (a potential system-state compromise); it does not request sudo, user creation, or direct editing of privileged system files elsewhere.