Last released Oct 20, 2026
A reliability layer for LLM API calls: retries, rate limiting, validation, tool calling, fallbacks, and structured logging.