notte-functions-forge

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

• Third-party content exposure detected (high risk: 0.75). During Phase 2/3 the skill uses notte-browser to explore a user-supplied target URL and then exports/executes session.scrape(...)/workflow code that reads scraped page content (untrusted outsider text) into the agent/LLM context for extraction, enabling indirect prompt injection.