If you've been running heavy workloads through Claude or Codex lately, you know the pain: those /status and /usage API calls either fail to load in your terminal or return garbage when you need them most. For power users burning through tokens like water, constantly refreshing the web dashboards manually isn't exactly a sustainable workflow.

The Problem With Native Token Tracking

The original poster on Hacker News—operating under jpthecat.com—put it simply: they don't trust the built-in status calls because 'they either fail to load in the CLI.' So what did they do? They started manually refreshing direct links to both Claude and Codex dashboards. That's not a solution, that's a band-aid on a hemorrhage. When you're running automated pipelines or debugging complex prompts, the last thing you need is another tab to babysit.

What This Little Cat Does

The tool—aptly named after its domain—is a lightweight utility that pulls your token usage data from both Anthropic's Claude and OpenAI's Codex directly. No more tab-switching. No more mysterious API failures killing your monitoring workflow. It gives you a unified view of where your tokens are going, which is exactly what the ecosystem has been missing for heavy users juggling multiple AI APIs simultaneously.

Why This Matters for the Ecosystem

Let's be real: token costs add up fast when you're doing serious development work. A tool that gives you accurate, at-a-glance visibility into consumption patterns across providers fills a genuine gap. The fact that this came from an individual developer solving their own pain point is peak hacker culture—scratching your own itch and sharing it with the community. That's how good tooling gets built outside corporate release cycles.

Key Takeaways

  • Solves real friction: unreliable /status and /usage API calls in CLI environments
  • Dual-provider support: tracks both Claude (Anthropic) and Codex (OpenAI)
  • Single dashboard approach eliminates manual tab management
  • Open-source project born from developer frustration with AI tooling gaps

The Bottom Line

This is the kind of utility that shouldn't need to exist—but until Anthropic and OpenAI fix their usage tracking reliability, someone had to build it. Grab it at jpthecat.com if you're tired of flying blind on your token burn rate.