New CLI proxy tool dramatically cuts LLM API costs by slashing token usage by up to 90% on routine developer commands

RTK introduces a command-line proxy that intercepts and optimizes requests to large language models

Tool achieves 60-90% reduction in token consumption for common development workflows

Project available on GitHub at rtk-ai/rtk for developers to integrate into their CLI tools

Targets cost optimization for teams frequently using LLM-powered development utilities