uppinote20/claude-dashboard

What it solves

Claude Code users often lack a unified way to monitor their resource usage, costs, and session status in real-time. This project provides a comprehensive status line plugin that aggregates usage data across multiple AI command-line interfaces (CLIs), allowing users to track token consumption, rate limits, and spending without leaving their terminal session.

How it works

The plugin integrates directly into the Claude Code status line via a modular widget system. It monitors and displays real-time data from Claude, Codex, Gemini, Antigravity, and z.ai CLIs. Users can configure the display through three preset modes (Compact, Normal, and Detailed) or create a custom layout by specifying which widgets to display on which line. It also includes a /claude-dashboard:check-usage command to compare available capacity across different AI providers and recommend the best one to use.

Who it’s for

Developers using Claude Code and other AI CLIs who want detailed visibility into their token usage, API costs, and rate limit countdowns to optimize their workflow and avoid unexpected interruptions.

Highlights

• Multi-Provider Monitoring: Unified tracking for Claude, Codex, Gemini, Antigravity, and z.ai.
• Modular Widget System: Dozens of customizable widgets for tracking model names, context window progress, session costs, and burn rates.
• Advanced Analytics: Includes cache hit rates, token breakdowns, and estimated time to rate limit depletion.
• Flexible Display Modes: Supports one-line, two-line, or six-line layouts with various themes and separators.
• Usage Recommendations: A dedicated command to check all AI CLI capacities and suggest the most available provider.