AnnualReports

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

• Third-party content exposure detected (high risk: 0.85). The REQUIRED workflow can run Tools/FetchReport.ts, which fetches arbitrary outsider-authored web page text at runtime via fetch(url) and then converts it to readable text (response.text() → extractTextFromHtml()), which is then written into a summary file and thus can be ingested into the agent’s LLM context as free-form content.