report-to-google-doc

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md:12-14 requires resolving a user-provided HTML report (possibly retrieved from a remote URL) and then the helper reads that HTML into readable text via parse_html() (scripts/report_to_google_doc/html_parser.py:648-653, 650-652), which is then used to generate skeleton.txt/manifest.json and ultimately fed into the agent’s LLM context as part of the agent handoff artifacts.