Hi Product Hunt! 👋 I built Shibaita.ai because Claude Code usage usually disappears into local logs and billing dashboards. I wanted a playful way to see how hard we are pushing it—and compare that with other developers—without uploading the work itself. The CLI reads local usage logs and submits only aggregated token counts. It does not send prompts, conversation text, source code, file paths, project names, environment variables, or credentials. Nothing is sent automatically: you run the submit command, preview the payload, and confirm it. Shibaita.ai turns those totals into daily and monthly rankings, public profiles, titles, and shareable cards. It is free and currently designed for Claude Code users. I’d love feedback on whether the setup feels clear and which ranking or privacy controls you would want next.

Shibaita.ai

A playful leaderboard for Claude Code usage

A playful leaderboard for Claude Code usage