fix: Decimal JSON serialization in chat_agent.py

feat: migrate chatbot tools to fully asynchronous direct-database queries to avoid loopback deadlocks

fix: resolve chatbot backend connection errors via dynamic port/url probing in tools

feat: multi-key rotation with graceful rate-limit handling for chat agent

fix: use fallback API key from environment variable for Groq rate limits

feat: enable general catalog queries in chatbot via fetch_all_products_metrics tool

Update fallback model from decommissioned llama3-8b-8192 to llama-3.1-8b-instant

Implement custom robust fallback using RunnableLambda in chat_agent.py

Add fallbacks to chatbot agent executor in chat_agent.py

Update backend routes and analytics endpoints to latest production version

Deploy FastAPI backend