agent-browser

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md core workflow and command reference explicitly show the agent navigating to arbitrary URLs (agent-browser open ), taking snapshots (agent-browser snapshot -i) and reading page content (get text/get html), which means it fetches and ingests untrusted public web pages (including via Browserbase) that can influence subsequent actions.