- Add in-memory ModelConfigCache (30s refresh, explicit invalidation)
replacing 2 SQLite queries per request (model lookup + cost override)
- Configure all 5 provider HTTP clients with proper timeouts (300s),
connection pooling (4 idle/host, 90s idle timeout), and TCP keepalive
- Move client_usage update to tokio::spawn in non-streaming path
- Use fast chars/4 heuristic for token estimation on large inputs (>1KB)
- Generate single UUID/timestamp per SSE stream instead of per chunk
- Add shared LazyLock<Client> for image fetching in multimodal module
- Add proxy overhead timing instrumentation for both request paths
- Fix test helper to include new model_config_cache field
- Set Grok to enabled: true by default.
- Updated AppState to include raw AppConfig.
- Refactored dashboard to show all supported providers, including their configuration and initialization status (online, disabled, or error).
- Added strict token validation against LLM_PROXY__SERVER__AUTH_TOKENS.
- Integrated 'reasoning_content' support into providers and server responses.
- Updated AppState to carry valid auth tokens for request-time validation.
