MathiasHC

Claude Code Token Dashboard — Development skill for Claude Code

Development community

Self-hosted dashboard for Claude Code token usage and API-equivalent cost.

How to install Claude Code Token Dashboard

This entry records only its repository, not the path inside it, so there is no exact command to give. Open MathiasHC/claude-code-token-dashboard and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Claude Code Token Dashboard does

Self-hosted dashboard for Claude Code token usage and API-equivalent cost. Reads your local transcripts, runs on any browser (even a first-gen iPad), zero dependencies.

Alternatives in Development

  • Sanitize — Detect and redact PII from text files — 15 categories (SSNs, credit cards, API keys, etc.), zero dependencies 3.1k ★
  • Claw Code Agent — Claw Code No Rust No TypeScript Only Python 543 ★
  • Claude Dashboard — Comprehensive status line plugin for Claude Code with context usage, API rate limits, and cost tracking 258 ★

README

Claude Code Token Dashboard

[![tests](https://github.com/MathiasHC/claude-code-token-dashboard/actions/workflows/tests.yml/badge.svg)](https://github.com/MathiasHC/claude-code-token-dashboard/actions/workflows/tests.yml)

A self-hosted dashboard for **Claude Code token usage and cost**. It reads the transcripts Claude Code already writes on your own machine, shows what you spent tokens on, and works out what that usage *would* have cost at Claude API list rates if you were paying per token instead of on a subscription.

It is built for an always-on display — a spare tablet, a small screen on a Raspberry Pi, a corner of a second monitor. The page carries no JavaScript and refreshes itself with a `` tag, so it runs on browsers far too old for anything modern.

Everything stays on your machine: no account, no telemetry, no network calls.

**No dependencies outside the Python standard library.**

![The dashboard showing token spend broken down by model, project, skill and source, with machine time and environmental footprint below](docs/screenshot.jpg)

The trophy ribbon on the right opens the all-time leaderboards over the top:

![Twelve all-time leaderboards in a sheet over the dashboard, each top three with gold, silver and bronze placings](docs/screenshot-leaderboards.jpg)

*Every figure above is synthetic demo data, not real usage. Reproduce it with `python3 tools/demo_page.py demo.html` and open the file — a way to see the layout before you have any history of your own.*

**What it shows, top to bottom:**

  • Live burn rate — $8.03/hr · idle 0m in the titlebar. Cost per hour of active time today; the only figure that changes between two refreshes.
  • Four windows — today, last 7 days, month to date, all time, each with a message count and a like-for-like delta (vs yesterday, vs the prior 7 days, vs the same point last month).
  • What it would cost vs what you pay — your api-equivalent spend against your actual plan, the resulting mult