Fix AI insight leaking thinking content + make insights on-demand instead of auto-generated

UX overhaul + Gradio migration + remove llama.cpp

fix: load LLM in background thread so Space stays healthy during 2.84 GB cold start

feat: browser-local engine, Zerodha dashboard, historical events, chatbot, per-user isolation

fix: serve static files, switch to FastAPI, add CRT terminal styling

Phase 2/3: Gradio Server backend, CRT frontend, engine, agents, mentor, tests, CI/CD