When your agent uses LiteLLM as a proxy, this command shows every API call made: model, prompt length, response time, and cost. Invaluable for debugging unexpected behavior and understanding what the agent is actually sending to the model.
Looking for more? Search all 7,657 commands — works offline, in English or Spanish, and fixes typos.