Reduce post-model chat response overhead

Restore Gemma primary and cut post-model API delay

Switch primary OpenRouter model to Meta Llama

Add multi-user mode to load test script

Improve Gemma failover and load-test realism

Fix cookie set compatibility in load script

Fix load test auth for cookie-based backend

Scale runtime for 20-user target load

Fix OpenRouter runtime error on HF

Unblock HF startup by de-risking Alembic boot

Prevent HF startup hangs during DB init

Add runtime request and release diagnostics

Harden OpenRouter routing and runtime observability

Fix chat model-error handling and user round auto-load

Harden model/provider config to stop persistent model errors

Add multi-model OpenRouter failover for model errors

Set DeepSeek as OpenRouter default with fallback

Tune response latency for concurrent user load

Improve latency path and 20-user scalability

Fix HF iframe embed and production auth over proxy

Fix Vercel login fetch failures for HF backend

Add Alembic recovery for existing/partial schemas

Fix HF startup migration loop on pre-existing schemas

Fix HF Space configuration metadata

Harden backend/frontend and clean repo artifacts

Fix SameSite cookie policy for cross-domain requests

Configure production environment: Aiven DB, HuggingFace Backend, Vercel ready

feat: add Docker containerization, orchestration, and async database configuration

fix: resolve all UPGRADED_DEEP_REPO_AUDIT critical issues

feat: async backend refactor + upgraded deep repo audit

updated files

updated to work with openrouter

deleted cache

configured backend

added API Routes

added requirments