010 GitHub stars
02Real-time cost and latency telemetry for each AI response.
03Multi-stage advisor pipeline with free and automatic premium fallback LLMs.
04Secure API key management using `.env` and `dotenv`.
05Zero-install deployment via `npx` for quick setup.
06Codex-native instructions for intelligent tool invocation by AI agents.