ClawMetry
ClawMetry is an open-source, zero-configuration real-time observability dashboard designed specifically for AI agent runtimes. It supports over 20 different agent runtimes, including Claude Code, OpenAI Codex, Cursor, Aider, Goose, OpenClaw, NVIDIA NemoClaw, NanoClaw, Hermes, and more. The tool automatically ingests telemetry data from these agents without requiring complex setup, providing immediate visibility into agent activities. The dashboard offers comprehensive monitoring capabilities, tracking live token usage, session data, tool calls, and cron job executions. It also provides cost analytics, enabling users to understand and optimize their AI agent spending. ClawMetry supports OpenTelemetry Protocol (OTLP) ingestion, making it compatible with existing observability stacks and allowing seamless integration into broader monitoring pipelines. As an open-source project hosted on GitHub, ClawMetry emphasizes ease of use and adaptability. Users can self-host the dashboard to gain full control over their data and customize the observability experience according to their needs. The zero-config nature means minimal friction for developers and teams looking to gain insights into their AI agent operations quickly.
- Category
- AI Agents
- Pricing
- Open Source
Problem solved
Monitoring and governing multiple AI agent runtimes is challenging due to the lack of unified visibility into sessions, token consumption, and costs. ClawMetry solves this by providing a zero-config, real-time dashboard that aggregates telemetry from various agent runtimes, enabling users to track performance, debug issues, and control expenses without complex setup.
Use cases
- Monitor token usage and costs across AI agent sessions
- Analyze tool call patterns and agent behavior
- Debug agent runtime issues with session tracking
- Govern AI agent usage in team environments
- Integrate with existing OTLP observability pipelines
- Track cron-based agent tasks and scheduled runs
Audience
- Developers
- ML Engineers
- Data Scientists
- Enterprise
- Startups
Pros
- Zero configuration setup
- Real-time observability
- Supports 20+ AI agent runtimes
- Open source and free
- Provides token and cost analytics
Cons
- Requires self-hosting
- May have limited documentation
- Early stage project with potential instability
- No managed cloud option
Official website: https://github.com/vivekchand/clawmetry
Canonical LambdaBase page: https://www.lambdabase.com/tools/clawmetry