browser-edge-cases

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

• Third-party content exposure detected (high risk: 0.85). The required workflow navigates to outsider-authored public web pages (e.g., https://www.linkedin.com/feed, https://x.com, https://example.com) and then uses browser tooling to extract readable DOM/snapshots via browser_html/browser_snapshot/browser_console, which can include attacker-controlled free text from those pages into the agent’s LLM context.