OpenCode plugin to query Z.ai GLM Coding Plan usage statistics with real-time quota monitoring, model usage tracking, and MCP tool usage.
- 📊 Query current quota limits (5-hour token cycle, monthly MCP usage)
- 🤖 View model usage statistics (24-hour rolling window)
- 🔧 View MCP tool usage (web_search, web_reader, etc.)
- 🌍 Supports both Global (api.z.ai) and CN (open.bigmodel.cn) platforms
- 🔐 Automatic credential discovery via OpenCode authentication context
- 📈 Visual progress bars for quota percentages (ASCII for terminal alignment)
- ⚡ Fail-fast error handling (no retry logic - user controls when to retry)
# Install plugin
npm install opencode-glm-quota
# Run installer to configure OpenCode
npx opencode-glm-quota installWhat the installer does:
- Copies
/glm_quotacommand to~/.config/opencode/command/glm_quota.md - Copies skill documentation to
~/.config/opencode/skills/glm-quota/SKILL.md - Automatically adds plugin to your OpenCode config
- Copies agent to
~/.config/opencode/agents/ - Supports
--forceflag to overwrite existing files
# Remove OpenCode integration files and config
npx opencode-glm-quota uninstall
# If installed globally
npx opencode-glm-quota uninstall --globalWhat the uninstaller does:
- Removes
/glm_quotacommand - Deletes
skills/glm-quota/SKILL.md - Removes plugin entry from OpenCode config
- Removes agent file and legacy config
- Runs
npm remove opencode-glm-quota(or--global)
npm install github:guyinwonder168/opencode-glm-quota# Clone and build
git clone https://github.com/guyinwonder168/opencode-glm-quota.git
cd opencode-glm-quota
npm install
npm run build
# Link for local testing
npm link
# Run the installer for local testing
node bin/install.jsOnce installed, simply run the plugin command in OpenCode:
/glm_quotaThe plugin will automatically detect your credentials (from OpenCode authentication) and display your usage statistics.
This plugin uses OpenCode's built-in authentication system. No manual configuration required.
Primary Method (Recommended):
# In OpenCode TUI
/connect
# Select: Z.AI Coding Plan or Z.AI
# Enter your API keyFallback (Development/Testing Only):
# For Global platform (api.z.ai)
export ZAI_API_KEY="your-api-key"
# For CN platform (open.bigmodel.cn)
export ZHIPU_API_KEY="your-api-key"After authentication, simply run:
/glm_quota╔══════════════════════════════════════════════════════════════════╗
║ ║
║ Z.ai GLM Coding Plan Usage Statistics ║
║ ║
╠══════════════════════════════════════════════════════════════════╣
║ Platform: Z.AI ║
║ Period: 2026-01-17 21:00:00 → 2026-01-18 20:59:59 ║
╠══════════════════════════════════════════════════════════════════╣
║ QUOTA LIMITS ║
╟──────────────────────────────────────────────────────────────────╢
║ Token usage(5 Hour) [###################--------] 40.5% ║
║ MCP usage(1 Month) [####-----------------------] 12.3% ║
║ Used: 123/1,000 ║
╠══════════════════════════════════════════════════════════════════╣
║ MODEL USAGE (24h) ║
╟──────────────────────────────────────────────────────────────────╢
║ Total Tokens (24h): 12,500,000 (31% of 5h limit) ║
║ 5h Window Usage: 40.5% of 40,000,000 ║
║ Total Calls: 1,234 ║
╠══════════════════════════════════════════════════════════════════╣
║ TOOL/MCP USAGE (24h) ║
╟──────────────────────────────────────────────────────────────────╢
║ Network Searches: 5,678 ║
║ Web Reads: 2,345 ║
║ ZRead Calls: 890 ║
╚══════════════════════════════════════════════════════════════════╝
The plugin uses fail-fast error handling. If any API request fails, it will display the error and stop (no automatic retries). This gives you full control over when to retry.
Example Error Output:
╔══════════════════════════════════════════════════════════════════╗
║ ║
║ ❌ Z.ai Credentials Not Found ║
║ ║
║ Please authenticate first: ║
║ ║
║ 1. Run /connect command in OpenCode TUI ║
║ 2. Select "Z.AI Coding Plan" or "Z.AI" ║
║ 3. Or "Zhipu" (for China region) ║
║ ║
║ For dev/testing, set environment: ║
║ - ZAI_API_KEY (global) ║
║ - ZHIPU_API_KEY (China) ║
║ ║
╚══════════════════════════════════════════════════════════════════╝
This plugin queries three Z.ai monitoring endpoints:
| Endpoint | Purpose | Query Params |
|---|---|---|
/api/monitor/usage/quota/limit |
Current quota percentages | None |
/api/monitor/usage/model-usage |
Model usage (24h window) | startTime, endTime |
/api/monitor/usage/tool-usage |
MCP tool usage (24h window) | startTime, endTime |
The plugin automatically detects the platform based on the provider ID used during authentication:
| Provider ID | Platform | API Base URL |
|---|---|---|
zai-coding-plan |
ZAI | https://api.z.ai |
zai |
ZAI | https://api.z.ai |
zhipu |
ZHIPU | https://open.bigmodel.cn |
The plugin discovers credentials in this order:
- OpenCode authentication context (managed by OpenCode) - PRIMARY
- Environment variable
ZAI_API_KEY(Global) orZHIPU_API_KEY(CN) - FALLBACK (dev/testing only)
Usage statistics are queried for a 24-hour rolling window:
- Start: Yesterday at current hour (e.g., 14:00:00)
- End: Today at current hour end (e.g., 14:59:59)
Critical: The plugin does NOT use "Bearer" prefix in the Authorization header. The token is passed directly:
Authorization: {token}
Accept-Language: en-US,en
Content-Type: application/json# Build TypeScript to JavaScript
npm run build
# Clean build artifacts
npm run clean
# Run all tests
npm run test
# Run test suite with coverage
npm run test:coverage
# Run specific test file
npm run test -- path/to/test.test.ts
# Watch mode during development
npm run test -- --watch
# Lint source code
npm run lint
# Prepare for npm publish
npm run prepublishOnlysrc/
index.ts # Main plugin entry point
api/
client.ts # HTTPS client with timeout and error handling
endpoints.ts # Platform-specific API endpoints
platforms.ts # Platform detection and naming
utils/
date-formatter.ts # Date/time formatting utilities
progress-bar.ts # ASCII progress bar rendering
time-window.ts # Rolling window calculation
integration/
agents/glm-quota-exec.md # Minimal executor agent (Markdown)
command/glm_quota.md # /glm_quota slash command
skills/glm-quota/SKILL.md # Skill documentation
bin/
install.js # Installation script
dist/ # Compiled JavaScript (generated)
tests/ # Test suite
package.json # Dependencies and scripts
tsconfig.json # TypeScript configuration
- Target: ES2022
- Module: NodeNext
- Strict mode enabled
- Always use type annotations for function returns
- Use
as constfor immutable constants
For detailed coding conventions, see AGENTS.md.
We welcome contributions! Please see CONTRIBUTING.md for guidelines.
MIT © 2026
- Adaptation from zai-org/zai-coding-plugins - This plugin is an OpenCode adaptation of the original Z.ai coding plugins repository
- Built for OpenCode
See CHANGELOG.md for version history and detailed changes.