🎯 Goal

Enable Jan to safely use a web browser via MCP (like BrowserMCP/ChromeMCP).
Let agents navigate, read pages, click, or type with clear user consent.

📖 Context

Jan already supports an MCP Hub and agent workflows.
Teams need in‑app browsing to fetch data and automate simple web tasks.

✅ Scope

Build a browser MCP client and APIs inside Jan.
Provide a “Use Browser” UI with run logs, approvals, and retries.
Support: open URL, extract content, click, type, screenshots. (upload/download?)

❌ Out of Scope

🛠 Deliverables

❓Open questions