

ClawMetry is a real-time observability dashboard specifically designed for OpenClaw AI agents. It provides comprehensive monitoring capabilities to help developers understand exactly what their AI agents are doing at any given moment.
The dashboard offers several key features including live flow visualization showing channels, gateway, models, tools, and nodes updating in real-time. It provides detailed sub-agent monitoring showing every step including what files agents are reading, what commands they're running, what tools they're calling, and their thought processes. The system includes token and cost tracking with metrics for tokens in/out, cache hits, response times, and cost per call. It also monitors system health including cron jobs, service uptime, disk usage, and active sub-agents.
ClawMetry integrates with OpenClaw's task system through Mission Control integration, allowing users to see which agents are assigned to which tasks. The dashboard maintains complete session history with timelines, tool calls, and costs for every session. It provides detailed cost breakdowns per session, per model, and per tool.
The product runs wherever OpenClaw runs, supporting macOS, Linux, Windows (including WSL), Raspberry Pi, and various cloud platforms like Digital Ocean, Hostinger, Hetzner, AWS, and GCP. Installation is designed to be simple with one-command installation via pip or platform-specific install scripts.
ClawMetry is targeted at developers working with OpenClaw AI agents who need visibility into their agent operations. It's particularly useful for monitoring complex agent workflows involving multiple sub-agents and understanding the costs and performance of AI operations.
admin
ClawMetry is designed for developers working with OpenClaw AI agents who need visibility into their agent operations. It's targeted at users who deploy AI agents that spawn sub-agents, burn tokens, and call tools, providing the monitoring capabilities that were previously missing. The tool serves developers running OpenClaw on various platforms including laptops, servers, Raspberry Pi, and cloud environments who want to understand exactly what their AI agents are doing in real-time.