agent-browser

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs the agent to open arbitrary public URLs (e.g., "agent-browser open " and the Google example "agent-browser open https://www.google.com") and to snapshot/get text/html from those pages, meaning the agent fetches and interprets untrusted third-party web content that can drive subsequent actions.