GLM Usage Status Plugin

A Claude Code plugin that displays GLM (ZHIPU/ZAI) coding plan usage statistics in the status bar.

Features

• Real-time Usage Tracking: Displays token and MCP usage percentages
• Color-coded Warnings:
  • Green (0-79%): Normal usage
  • Yellow (80-94%): High usage
  • Red (95-100%): Critical usage
• Smart Caching: 5-minute cache to reduce API calls
• Automatic Platform Detection: Supports both ZAI and ZHIPU platforms
• Graceful Degradation: Silently fails on API errors

Installation

Method 1: Via npm (Recommended ⭐)

The easiest way to install and manage the plugin:

npm install -g glm-plan-usage

This installs the plugin globally and makes it available as glm-plan-usage in your terminal.

Initialize Configuration:

glm-plan-usage --init

Verify Installation:

glm-plan-usage --help

Supported Platforms:

• ✅ Linux (x64, arm64)
• ✅ macOS (Intel, Apple Silicon)
• ✅ Windows (x64)

Method 2: Manual Installation from Source

Build from source if you prefer or need customization:

1. Build

cd glm-plan-usage
cargo build --release

2. Install

mkdir -p ~/.claude/glm-plan-usage
cp target/release/glm-plan-usage ~/.claude/glm-plan-usage/
chmod +x ~/.claude/glm-plan-usage/glm-plan-usage

3. Initialize Configuration

~/.claude/glm-plan-usage/glm-plan-usage --init

This creates a default config at ~/.claude/glm-plan-usage/config.toml:

[style]
mode = "plain"
separator = " | "

[[segments]]
id = "glm_usage"
enabled = true

[segments.colors]
text = { c256 = 109 }  # Default green

[segments.styles]
text_bold = true

[api]
timeout_ms = 5000
retry_attempts = 2

[cache]
enabled = true
ttl_seconds = 300

4. Configure Claude Code

If installed via npm:

Edit ~/.config/claude-code/settings.json:

{
  "statusLine": {
    "type": "command",
    "command": "glm-plan-usage",
    "padding": 0
  }
}

If installed manually:

{
  "statusLine": {
    "type": "command",
    "command": "~/.claude/glm-plan-usage/glm-plan-usage",
    "padding": 0
  }
}

5. Set Environment Variables

Add to your ~/.bashrc or ~/.zshrc:

export ANTHROPIC_AUTH_TOKEN="your-token-here"
export ANTHROPIC_BASE_URL="https://open.bigmodel.cn/api/anthropic"

Usage

The plugin automatically displays usage in the format:

T:42% M:15%

Where:

• T:42% = Token usage (5-hour window)
• M:15% = MCP usage (30-day window)

Using the npm-installed Version

Initialize config:

glm-plan-usage --init

Test the plugin:

# Set environment variables
export ANTHROPIC_AUTH_TOKEN="your-token"
export ANTHROPIC_BASE_URL="https://open.bigmodel.cn/api/anthropic"

# Test with sample input
echo '{"model":{"id":"test"}}' | glm-plan-usage

Update config manually (if needed):

# Edit config directly
nano ~/.claude/glm-plan-usage/config.toml

# Or reinitialize
glm-plan-usage --init

CLI Options

glm-plan-usage [OPTIONS]

Options:
  --init        Initialize configuration file
  --verbose     Enable verbose output
  --no-cache    Disable cache
  -h, --help    Print help

Configuration

Style Settings

[style]
mode = "plain"  # plain, nerd_font, powerline
separator = " | "

Segment Colors

[[segments]]
id = "glm_usage"
enabled = true

[segments.colors]
text = { c256 = 109 }  # 256-color palette
# OR RGB:
# text = { r = 34, g = 197, b = 94 }

API Settings

[api]
timeout_ms = 5000      # HTTP timeout
retry_attempts = 2     # Number of retries

Cache Settings

[cache]
enabled = true
ttl_seconds = 300      # 5 minutes

API Endpoints

The plugin queries the GLM API:

• Quota Limits: GET /api/monitor/usage/quota/limit
• Token Usage: GET /api/monitor/usage/model-usage
• MCP Usage: GET /api/monitor/usage/tool-usage

Error Handling

The plugin gracefully handles errors:

• Missing environment variables → No output
• API timeout → Uses cached data (if available)
• API errors → Uses cached data (if available)
• No cache → Silent failure

Use --verbose to debug issues:

# If installed via npm
glm-plan-usage --verbose < input.json

# If installed manually
~/.claude/glm-plan-usage/glm-plan-usage --verbose < input.json

Updating

Via npm

# Update to the latest version
npm update -g glm-plan-usage

# Or reinstall
npm install -g glm-plan-usage@latest

Manual Installation

# Rebuild from source
cd glm-plan-usage
cargo build --release

# Reinstall
cp target/release/glm-plan-usage ~/.claude/glm-plan-usage/

Development

Project Structure

glm-plan-usage/
├── Cargo.toml
├── README.md
└── src/
    ├── main.rs              # Entry point
    ├── cli.rs               # CLI parsing
    ├── lib.rs               # Library root
    ├── config/
    │   ├── mod.rs
    │   ├── types.rs         # Config types
    │   └── loader.rs        # Config loading
    ├── core/
    │   ├── mod.rs
    │   ├── statusline.rs    # ANSI rendering
    │   └── segments/
    │       ├── mod.rs
    │       └── glm_usage.rs # GLM usage segment
    └── api/
        ├── mod.rs
        ├── client.rs        # HTTP client
        └── types.rs         # API types

Build Release

cargo build --release

The binary is stripped and optimized for size.

License

MIT