browse

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

• Third-party content exposure detected (high risk: 1.00). This skill's SKILL.md instructs the agent to open arbitrary web pages with "agent-browser open " and to inspect/get page text/html and use snapshots, clicks, and form fills (e.g., "agent-browser get text body", "agent-browser snapshot -i", "agent-browser click ..."), meaning it fetches and interprets untrusted public web content that can influence subsequent actions.