# Coding Agent Usage Tracker

A single command-line tool that checks remaining quota, rate limits and cost across more than 16 AI coding assistant and LLM providers, instead of checking each provider's own dashboard.

caut fetches usage data from more than 16 LLM providers through one command, printing human-readable tables in the terminal by default, or structured JSON or Markdown meant for an AI agent to consume, and tracks session, weekly and model-tier rate limits alongside real-time credit balances, provider outage status, and cost tracked locally by scanning usage logs rather than exporting a CSV from each provider by hand.

Each of the 16 providers is implemented as a descriptor carrying its own metadata, branding and fetch strategy, so adding a new provider means implementing one trait rather than touching the tool's core logic. It is installed from source with Cargo, runs on macOS, Linux and Windows, and includes its own AGENTS.md file documenting toolchain requirements, coding conventions and testing expectations for AI agents that contribute to the codebase. It is written in Rust.

Website: https://github.com/Dicklesworthstone/coding_agent_usage_tracker
