Local CLI that pinpoints why your LLM bill spiked — runaway agent loops, silent model swaps, bloated prompts, costly endpoints. Logs in, report out.
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).