百科.dev
全部条目AI 编程趋势榜开源项目技术资讯提交条目
登录
< 返回工具列表
C

Claude-Code-Usage-Monitor

> 编程语言
开源

实时 Claude 代码使用监控,包括预测和警告

8.6K stars0 点赞0 次浏览
访问官网GitHub

工具介绍

实时 Claude 代码使用监控,包括预测和警告

Claude Code Usage Monitor

A privacy-first Claude Usage-Ops companion for Claude Code. It combines a Rich live terminal monitor with official statusline rate_limits, machine-readable state/export output, provenance labels, forecasting, and an opt-in local usage warehouse.


Table of Contents

  • ✨ Key Features
  • Installation
    • ⚡ Modern Installation with uv (Recommended)
    • Installation with pip
    • ️ Other Package Managers
  • Usage
    • Get Help
    • Basic Usage
    • Configuration Options
    • Available Plans
  • Please Help Test This Release!
  • ✨ Features & How It Works
    • Current Features
    • Understanding Claude Sessions
    • Token Limits by Plan
    • Smart Detection Features
  • Usage Examples
    • Common Scenarios
    • Best Practices
  • Development Installation
  • Troubleshooting
    • Installation Issues
    • Runtime Issues
  • Contact
  • Related Tools
  • Additional Documentation
  • License
  • Contributors
  • Acknowledgments

✨ Key Features

v4.0.0 Major Update - Usage Ops Companion

  • ** Official-limit trust layer** - --statusline captures Claude Code's official rate_limits; stale or expired captures fall back to labeled local estimates.
  • ** Machine-readable protocol** - --once, --compact, and --write-state all use one versioned snapshot builder with automation exit codes.
  • ️ Provenance labels - exported and displayed numbers distinguish official, local_estimate, experimental, and unknown confidence.
  • ** Persistent usage warehouse** - opt-in local history survives Claude's 30-day cleanup with project/model/day dimensions and CSV/JSON reports.
  • ** Forecasting and pace** - reset-aware pace, date-context forecasts, official-only weekly percentages, and limit-hit freeze behavior.
  • ** Multi-source input** - --data-paths, CLAUDE_CONFIG_DIR, and WSL discovery can scan multiple directories without merging unrelated accounts into one 5-hour window.
  • ️ Rich UI parity - live Rich output, rich one-shot output, compact output, state files, and exports use the same snapshot contract.
  • ** External companion boundary** - GUIs, trays, provider adapters, and status bars should consume --write-state or --once --output json.
  • ** Regression coverage** - the non-integration suite now covers the trust layer, state protocol, warehouse, reports, title updates, multi-source paths, and timezone edge cases.

Default Custom Plan

The Custom plan is now the default option, specifically designed for 5-hour Claude Code sessions. It monitors three critical metrics:

  • Token usage - Tracks your token consumption
  • Messages usage - Monitors message count
  • Cost usage - The most important metric for long sessions

The Custom plan automatically adapts to your usage patterns by analyzing all your sessions from the last 192 hours (8 days) and calculating personalized limits based on your actual usage. This ensures accurate predictions and warnings tailored to your specific workflow.

Installation

⚡ Modern Installation with uv (Recommended)

Why uv is the best choice:

  • ✅ Creates isolated environments automatically (no system conflicts)
  • ✅ No Python version issues
  • ✅ No "externally-managed-environment" errors
  • ✅ Easy updates and uninstallation
  • ✅ Works on all platforms

The fastest and easiest way to install and use the monitor:

Install from PyPI

# Install directly from PyPI with uv (easiest)
uv tool install claude-monitor

# Run from anywhere
claude-monitor  # or cmonitor, ccmonitor for short

Install from Source

# Clone and install from source
git clone https://github.com/Maciek-roboblog/Claude-Code-Usage-Monitor.git
cd Claude-Code-Usage-Monitor
uv tool install .

# Run from anywhere
claude-monitor

First-time uv users

If you don't have uv installed yet, get it with one command:

# On Linux/macOS:
curl -LsSf https://astral.sh/uv/install.sh | sh

# On Windows:
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

# After installation, restart your terminal

Installation with pip

# Install from PyPI
pip install claude-monitor

# If claude-monitor command is not found, add ~/.local/bin to PATH:
echo 'export PATH="$HOME/.local/bin:$PATH"' >> ~/.bashrc
source ~/.bashrc  # or restart your terminal

# Run from anywhere
claude-monitor  # or cmonitor, ccmonitor for short

⚠️ PATH Setup: If you see WARNING: The script claude-monitor is installed in '/home/username/.local/bin' which is not on PATH, follow the export PATH command above.

⚠️ Important: On modern Linux distributions (Ubuntu 23.04+, Debian 12+, Fedora 38+), you may encounter an "externally-managed-environment" error. Instead of using --break-system-packages, we strongly recommend:

  1. Use uv instead (see above) - it's safer and easier
  2. Use a virtual environment - python3 -m venv myenv && source myenv/bin/activate
  3. Use pipx - pipx install claude-monitor

See the Troubleshooting section for detailed solutions.

️ Other Package Managers

pipx (Isolated Environments)

# Install with pipx
pipx install claude-monitor

# Run from anywhere
claude-monitor  # or claude-code-monitor, cmonitor, ccmonitor, ccm for short

conda/mamba

# Install with pip in conda environment
pip install claude-monitor

# Run from anywhere
claude-monitor  # or cmonitor, ccmonitor for short

Usage

Get Help

# Show help information
claude-monitor --help

Available Command-Line Parameters

Parameter Type Default Description --plan string custom Plan type: pro, max5, max20, team, or custom --custom-limit-tokens int None Token limit for custom plan (must be > 0) --view string realtime View type: realtime, daily, monthly, session, entries, sessions, or burn-rate --output string rich Output format: rich, json, text, or csv --once flag False Measure once, print a snapshot, and exit --compact flag False Single-line compact output for status bars --write-state flag False Write the snapshot to a state file for external tools --state-file path None State file path for --write-state --statusline flag False Run as a Claude Code statusline hook and capture official rate_limits --api flag False Enable the opt-in experimental Anthropic OAuth usage API --data-paths list [] Claude data directories to scan; repeat or comma-separate values --warehouse flag False Persist usage entries to the opt-in local warehouse --warehouse-file path None Usage warehouse file path --warehouse-retention-days int 365 Days of warehouse records to retain --timezone string auto Timezone (auto-detected). Examples: UTC, America/New_York, Europe/London --time-format string auto Time format: 12h, 24h, or auto --theme string auto Display theme: light, dark, classic, or auto --refresh-rate int 10 Data refresh rate in seconds (1-60) --refresh-per-second float 0.75 Display refresh rate in Hz (0.1-20.0) --reset-hour int None Daily reset hour (0-23) --date-format string None Date format for daily/monthly table periods --abbreviate-tokens flag False Abbreviate token counts in table views --sparklines flag False Show opt-in sparklines in table views --filter-models string all Use all models, or anthropic to exclude routed non-Claude models --set-terminal-title flag False Set the terminal title from the usage snapshot --title-format string "{pct}% {plan}" Terminal title template using pct, plan, used, limit, cost, reset --hide-model-distribution flag False Hide the model distribution bar --no-header flag False Hide the header banner --no-emoji flag False Render plain output without emoji --log-level string INFO Logging level: DEBUG, INFO, WARNING, ERROR, CRITICAL --log-file path None Log file path --debug flag False Enable debug logging --version, -v flag False Show version information --clear flag False Clear saved configuration

Plan Options

Plan Token Limit Cost Limit Description pro 19,000 $18.00 Claude Pro subscription max5 88,000 $35.00 Claude Max5 subscription max20 220,000 $140.00 Claude Max20 subscription team estimate unverified Team label; prefer official statusline or --plan custom custom P90-based (default) $50.00 Auto-detection with ML analysis

Command Aliases

The tool can be invoked using any of these commands:

  • claude-monitor (primary)
  • claude-code-monitor (full name)
  • cmonitor (short)
  • ccmonitor (short alternative)
  • ccm (shortest)

Save Flags Feature

The monitor automatically saves your preferences to avoid re-specifying them on each run:

What Gets Saved:

  • View type (--view)
  • Plan (--plan)
  • Theme preferences (--theme)
  • Timezone settings (--timezone)
  • Time format (--time-format)
  • Refresh rates (--refresh-rate, --refresh-per-second)
  • Reset hour (--reset-hour)
  • Custom token limits (--custom-limit-tokens)

Configuration Location: ~/.claude-monitor/last_used.json

Usage Examples:

# First run - specify preferences
claude-monitor --plan pro --theme dark --timezone "America/New_York"

# Subsequent runs - preferences automatically restored
claude-monitor --plan pro

# Override saved settings for this session
claude-monitor --plan pro --theme light

# Clear all saved preferences
claude-monitor --clear

Key Features:

  • ✅ Automatic parameter persistence between sessions
  • ✅ CLI arguments always override saved settings
  • ✅ Atomic file operations prevent corruption
  • ✅ Graceful fallback if config files are damaged
  • ✅ Plan is saved and restored between sessions (override any time with --plan)
  • ✅ An explicitly chosen theme is kept and not overwritten by auto-detection

Basic Usage

With uv tool installation (Recommended)

# Default (Custom plan with auto-detection)
claude-monitor

# Alternative commands
claude-code-monitor  # Full descriptive name
cmonitor             # Short alias
ccmonitor            # Short alternative
ccm                  # Shortest alias

# Exit the monitor
# Press Ctrl+C to gracefully exit

Development mode

If running from source, use python -m claude_monitor from the src/ directory.

Machine-readable Usage Protocol

Use these surfaces when another tool needs current usage without scraping the Rich TUI:

# One-shot JSON snapshot with source/confidence/provenance fields
claude-monitor --once --output json

# Compact one-line status output
claude-monitor --once --compact

# Keep an atomically-written state file up to date for status bars and dashboards
claude-monitor --write-state --state-file ~/.claude-monitor/state/latest.json

# Install as a Claude Code statusline hook to capture official rate_limits
claude-monitor --statusline

--once exits with automation-friendly code

GitHub Issues· 0 开放

在 GitHub 查看全部

暂无开放 Issues,或尚未同步最近议题。

核心特点

  • •✨ Key Features
  • •Installation
  • •⚡ Modern Installation with uv (Recommended)
  • •Installation with pip
  • •️ Other Package Managers
  • •Get Help
  • •Basic Usage
  • •Configuration Options
  • •Available Plans
  • •Please Help Test This Release!

> 标签

Pythonaianalyticsclaudeclaude-code

暂无评论,来聊聊你的看法吧

> 工具信息

发布日期2026年8月1日
最后更新2026年9月17日
分类编程语言
定价开源

> 相关工具

T
TypeScript
JavaScript 的超集,为前端与全栈提供静态类型
P
Python
通用编程语言,广泛用于 Web、数据与 AI
G
Go
Google 推出的简洁高效系统语言