momentic-test

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

• Third-party content exposure detected (high risk: 1.00). The skill's core workflow instructs the agent to NAVIGATE to the test's baseUrl and to call momentic_preview_step / momentic_get_session_state to read UI-state snapshots and linked page/artifact content (e.g., "momentic_preview_step NAVIGATE to baseUrl" and "use your Read tool on the linked path"), which means the agent will fetch and interpret arbitrary external web pages provided as baseUrl or discovered during sessions, allowing untrusted third-party page content to influence step construction and tool use.