Show HN: I built a GitHub-style contribution graph for Claude Code usage

I've been using Claude Code a bunch lately and wanted to see my usage patterns visualized like GitHub's contribution graph (mostly out of curiosity).

So I built ccheatmap - a tiny CLI tool that reads your local Claude Code usage data and renders it as a heatmap right in your terminal.

npx ccheatmap

It just reads the JSON files Claude Code already writes to ~/.config/claude/projects/ and shows you:

- Daily session counts (or token usage, or interactions) - Weekly patterns (turns out I use Claude most on Tuesdays ) - Total stats for the period - Nice color gradients to represent activity

The whole thing is ~300 lines of TypeScript. Nothing fancy - just parsing timestamps, bucketing by day, and rendering colored Unicode blocks.

What surprised me: weekend usage often spiked way higher than weekdays. Turns out I tend to vibe code more on side projects when the stakes are lower.

Repo: https://github.com/viveknair/ccheatmap

Would love to hear if others track their AI tool usage differently, or if there are other metrics that would be interesting to visualize. Also curious if anyone else has noticed usage patterns they didn't expect.

I was inspired by ryoppippi's ccusage tool which has cost tracking and detailed usage analytics - definitely check that out if you need deeper insights.