Switch Groq model to openai/gpt-oss-120b, bump max_tokens to 4096

Update default Groq model to llama-3.3-70b-versatile

Implement infra hardening: streaming, refresh tokens, structlog, Sentry, Redis

Context-aware chat with history autofill

Production-grade: guest mode, daily limits, modal auth, inline confirms

Polish UI, clean up backend, improve code quality

updated error handling in utils

cleaned app and utils

resolved env issues in utils

first working version with groq api, and added a file to check available groq models

added base infrastructure and gitignored env vars