mudrii/openclaw-dashboard

What it solves

OpenClaw Dashboard provides a centralized, local command center for managing complex OpenClaw AI agent deployments. It eliminates the need to manually parse log files, run CLI commands, or track costs across multiple models and agents, providing an at-a-glance overview of system health, spending, and agent activity.

How it works

The project is a zero-dependency application consisting of a Go binary backend and a pure HTML/CSS/JS frontend. It collects data from an OpenClaw installation via a CLI adapter and generates a data.json file. The frontend retrieves this data through a "stale-while-revalidate" API, ensuring instant page loads while triggering background refreshes to keep information current. It also integrates with the OpenClaw gateway's HTTP endpoint to allow natural language queries about the system's state.

Who it’s for

Users running multiple OpenClaw agents, sub-agents, and scheduled cron jobs across various platforms (Telegram, WhatsApp, Slack, Discord) who need a visual way to monitor costs, performance, and runtime health.

Highlights

• Comprehensive Monitoring: Tracks live CPU/RAM/disk usage, gateway health, and detailed cost breakdowns per model.
• Agent Observability: Monitors active sessions, sub-agent task progress, and cron job status (including "flapping" unstable jobs).
• AI-Powered Insights: Includes a built-in chat interface to ask questions about costs, sessions, and configuration using the OpenClaw gateway.
• Zero-Dependency Architecture: Runs as a single binary with an embedded frontend, requiring no external databases or cloud services.
• Customizable UI: Features a glass morphism design with six built-in themes and support for custom CSS color variables.