openclaw-dashboard

Real-time monitoring dashboard for OpenClaw AI agents.

Features

• 📊 Real-time monitoring — Agent status, token usage, and costs update live via file watching
• 💰 Cost insights — Subscription utilization, per-model/per-agent cost breakdown
• 📈 Charts — Daily cost and token usage with cumulative trends (Chart.js)
• 🔄 Auto-detection — Reads openclaw.json to discover agents, models, and subscriptions
• 🌙 Dark theme — Clean, modern dark UI
• 🔌 Plugin architecture — Runs as an OpenClaw plugin (HTTP handler, RPC, CLI, slash command)

Quick Start (30 seconds)

# 1. Install
openclaw plugins install openclaw-dashboard

# 2. Restart gateway to load the plugin
openclaw gateway restart

# 3. Open in browser
open http://localhost:3578/dashboard

You'll see a dark-themed dashboard with:

• Summary cards — Monthly subscription cost, estimated API value, total tokens, active days
• Subscription utilization bar — How much of your subscription you're actually using
• Agent breakdown — Per-agent cost and token stats
• Daily charts — Stacked bar charts for cost and tokens by agent/model
• Model table — Detailed per-model usage with pricing tier info
• Active sessions — Live session list with status and token counts

Note: The dashboard auto-detects agents and subscriptions from your openclaw.json. No extra configuration needed.

Installation

openclaw plugins install openclaw-dashboard
openclaw gateway restart

Usage

Web Dashboard

Open http://localhost:<gateway-port>/dashboard in your browser.

CLI

openclaw dashboard              # Text summary
openclaw dashboard --json       # Full JSON output
openclaw dashboard --from 2026-01-01  # Date range filter

Slash Command

Send /dashboard in any connected chat channel.

RPC

openclaw rpc dashboard.status
openclaw rpc dashboard.snapshot '{"from":"2026-01-01"}'

Configuration

In your openclaw.json:

{
  "plugins": {
    "entries": {
      "openclaw-dashboard": {
        "enabled": true,
        "config": {
          "basePath": "/dashboard",
          "refreshIntervalMs": 10000,
          "theme": "dark"
        }
      }
    }
  }
}

| Option | Default | Description | |--------|---------|-------------| | basePath | /dashboard | URL path for the dashboard UI | | refreshIntervalMs | 10000 | Data collection interval (ms) | | theme | dark | Color theme (dark or light) |

How It Works

1. Data Collection: Parses JSONL session files in ~/.openclaw/agents/*/sessions/ for historical cost/token data
2. Active Sessions: Calls openclaw sessions --json --active 120 for live session info
3. Auto-Detection: Reads openclaw.json to discover agents, models, auth profiles, and subscriptions
4. File Watching: Uses fs.watch on session directories for near-instant updates
5. Serving: Registers HTTP routes on the Gateway — no separate server needed

What It Tracks

• Per-agent: Cost, tokens (input/output/cache), model usage
• Per-model: Pricing tier, subscription vs pay-per-use classification
• Subscriptions: Auto-detects from auth profiles (Anthropic Max, OpenAI Plus, Google AI Pro)
• Daily trends: Stacked bar charts with cumulative lines
• Active sessions: Live session status, subagents, cron jobs

License

MIT