fix: add pool_pre_ping and pool_recycle to prevent stale Neon connection errors

fix(agent): handle non-dict LLM response gracefully and delay setting task to FAILED during Celery retries

feat(api): filter cases by authenticated user and protect endpoints

feat(api): add endpoint to list past cases

revert LLMService back to official Google GenAI SDK

refactor LLMService to route Gemini requests through OpenRouter

integrate ResearchAgent loop and update_hypothesis database updates in Celery worker

feat: add prometheus metric tracking for token counts, agent latency, and failure rates

chore: upgrade default gemini model to 2.5-flash and remove hardcoded models

feat: implement google gemini llm helper service with structured json configuration

feat: implement shared context manager for Postgres mapping and context state updates

fix: resolve db sslmode string space and foreign key table naming

feat: created neon db connection using sqlalchemy

feat: added files and folders