Claumon

See where your Claude Code limits are headed before you hit them - and manage everything Claude Code leaves on your machine.

Works with any claude /login plan: Pro, Max, or a Team/Enterprise seat. Forecasts come from a published, benchmarked statistical model, not a burn-rate guess. Single binary, zero config, one browser tab. Runs on macOS, Linux, and Windows.

Why claumon?

Anthropic's usage analytics dashboard is for Team and Enterprise org admins; it is not available to individual Pro or Max plans, and even for admins it is org-level analytics, not a live view of your own machine. What you get personally (/usage, the claude.ai usage page) shows where you stand right now - but no history, no per-session costs, and no idea where you're heading before the limit hits. claumon is that missing personal dashboard, running locally, whatever plan you're on.

What sets it apart:

- Forecasts you can check. Projected utilization at reset with an 80% credible interval and ETA-to-threshold, from an empirical-Bayes model refit daily on your own usage history. The model is published (MODEL.pdf) and benchmarked out-of-sample with proper scoring rules - not a burn-rate extrapolation or a percentile heuristic.
- Live rate-limit gauges. Session, weekly, and per-model utilization straight from the Claude OAuth usage API - measured, not estimated from logs.
- Your whole ~/.claude footprint in one place. Usage trackers stop at tokens and cost; claumon also manages what Claude Code accumulates on disk - per-session cost breakdowns, historical trends in SQLite, running-process control, and a memory-file browser with health scores, a relationship graph, and recoverable deletion.

Everything updates in real time via SSE, and daily aggregates persist in SQLite so you can track usage over…