cnki-export

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

• Third-party content exposure detected (high risk: 0.75). The required workflow uses browser action=act kind=evaluate to fetch and parse CNKI page content (e.g., document.body.innerText and search-result DOM like input.cbItem/rows) at runtime, which is outsider-authored free text from a public web source and can be injected into the LLM context via the tool’s returned JSON.