google-sheets-plg

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow reads outsider-provided Google Sheets content by making Sheets API reads (e.g., record/cell tools like createRow/createRows call readHeaders which fetches .../values/{sheet}!1:1, and listRows/getValues likewise fetch A1/ranges) after the agent selects a spreadsheet/worksheet based on user input, so free text inside the target spreadsheet becomes LLM-ingestible at runtime.