comet-ml/opik-openclaw

🦞 Official plugin for OpenClaw that exports agent traces to Opik. See and monitor agent behaviour, cost, tokens, errors and more.

What it solves

It provides observability and monitoring for OpenClaw agent runs by exporting detailed execution traces to the Opik platform. This allows developers to track exactly how an agent is behaving, including its interactions with LLMs and tools.

How it works

The plugin integrates directly into the OpenClaw Gateway process. It listens for specific OpenClaw events—such as LLM inputs/outputs, tool calls, and sub-agent spawning—and maps them to Opik entities (traces and spans). This captures request/response data, tool inputs/outputs, errors, and cost/usage metadata.

Who it’s for

Developers using OpenClaw to build AI agents who need to debug, monitor, and evaluate their agent's performance and execution flow using Opik.

Highlights

  • Detailed Tracing: Captures LLM request/response spans, sub-agent lifecycles, and tool call details.
  • Metadata Tracking: Records usage and cost metadata for runs.
  • Easy Setup: Includes a CLI-based interactive setup wizard (openclaw opik configure).
  • Native Integration: Operates as a plugin within the OpenClaw ecosystem using native hooks.

Related

  • Project
    comet-ml/opik

    An open-source LLM observability and evaluation platform for tracing AI agents, managing prompts, and monitoring production LLM applications.

  • Project
    lmnr-ai/lmnr

    An open-source observability platform for AI agents that provides tracing, monitoring, and evaluation tools to debug and optimize agentic workflows.

  • Project
    mudrii/openclaw-dashboard

    A zero-dependency local command center for OpenClaw AI agents that provides real-time monitoring of costs, gateway health, and agent activity.

  • Project
    simple10/agents-observe

    A real-time observability dashboard for Claude Code and Codex agents that provides visibility into subagent hierarchies, tool calls, and token usage.