OpenClaw Dashboard

OpenClaw Dashboard is a local monitoring dashboard for the OpenClaw AI agent framework, built with a zero-dependency architecture (pure HTML/CSS/JS frontend ~1160 lines, Python stdlib backend ~400 lines).

Core Features#

11 Dashboard Panels:

• Header Bar: Bot name, online/offline status, auto-refresh countdown, theme selector
• Alerts Banner: Smart alerts (high cost, cron failures, high context usage, gateway offline)
• System Health: Gateway status, PID, uptime, memory, compression mode, active sessions
• Cost Cards: Today's cost, total cost, monthly estimate, cost distribution donut chart
• Token Usage & Cost: Token and cost breakdown by model (7d/30d/all-time tabs)
• Cron Jobs: Scheduled task status, schedule, last/next run time, model
• Active Sessions: Recent sessions (model, type tags, context %, tokens)
• Sub-Agent Activity: Sub-agent run records (cost, duration, status, token breakdown)
• Charts & Trends: Cost trend line charts, model cost bar charts (pure SVG)
• Bottom Row: Available models grid, skills list, Git log
• AI Chat: Natural language queries for dashboard data

Key Features:

• 6 built-in themes (3 dark + 3 light, 19 CSS variables for customization)
• Smart refresh: On-demand + configurable auto-refresh (default 30s debounce)
• 5-level model parsing chain for accurate model names
• Local-first, no cloud dependencies

Installation#

Requirements: Python 3.6+, OpenClaw installed at ~/.openclaw, modern browser, macOS 10.15+ / Linux (Ubuntu 18.04+, Debian 10+)

One-line Install:

curl -fsSL https://raw.githubusercontent.com/mudrii/openclaw-dashboard/main/install.sh | bash

Manual Install:

git clone https://github.com/mudrii/openclaw-dashboard.git ~/.openclaw/dashboard
cd ~/.openclaw/dashboard
cp examples/config.minimal.json config.json
python3 server.py &

Configuration#

Key config.json settings:

• bot.name / bot.emoji: Bot display name and icon
• theme.preset: Theme preset (nord, etc.)
• refresh.intervalSeconds: Refresh interval
• server.port / server.host: Server port and bind address
• ai.enabled / ai.gatewayPort: AI Chat feature config
• alerts: Alert thresholds (daily cost, context usage, memory)

Config priority: CLI flags > Environment variables > config.json > Defaults

Architecture#

Data Pipeline: Browser → server.py → refresh.sh → data.json (atomic write)

API Endpoints:

• GET /api/refresh: Refresh and return dashboard data JSON
• POST /api/chat: AI natural language query

Data Sources: openclaw.json, agents//sessions/.jsonl, cron/jobs.json, process table

Security Notes#

• Default bind 127.0.0.1 (localhost only)
• LAN mode has no built-in authentication
• No HTTPS, requires reverse proxy for TLS
• data.json contains sensitive data