logfire-observability

name	logfire-observability
description	Full OpenClaw observability in Pydantic Logfire — agent traces, tool calls, metrics, and logs via OpenTelemetry
homepage	https://github.com/rita-aga/openclaw-logfire-observability
metadata	{"openclaw":{"emoji":"🔥","requires":{"env":["LOGFIRE_TOKEN"]},"os":["darwin","linux","win32"],"primaryEnv":"LOGFIRE_TOKEN"}}

Logfire Observability

Sends OpenClaw agent traces, tool calls, and messages to Pydantic Logfire via OpenTelemetry.

Install the plugin

openclaw plugins install openclaw-logfire-observability

Configure

Add to your openclaw.json:

{
  "diagnostics": {
    "enabled": true,
    "otel": {
      "enabled": true,
      "endpoint": "https://logfire-us.pydantic.dev",
      "headers": {
        "Authorization": "Bearer pylf_v1_us_YOUR_TOKEN_HERE"
      },
      "serviceName": "openclaw",
      "traces": true,
      "metrics": true,
      "logs": true
    }
  },
  "plugins": {
    "entries": {
      "openclaw-logfire-observability": {
        "enabled": true,
        "config": {
          "logfireToken": "pylf_v1_us_YOUR_TOKEN_HERE"
        }
      },
      "diagnostics-otel": {
        "enabled": true
      }
    }
  }
}

Replace YOUR_TOKEN_HERE with your Logfire write token (from Settings > Write Tokens in your Logfire project).

Then restart OpenClaw.

What you get

• Agent traces with parent-child nesting (agent.run → tool.* spans)
• Metrics: token usage, cost, duration, queue health, session state
• Logs: full OpenClaw log forwarding
• Diagnostic traces: model usage, webhooks, message processing

See the full README for architecture details, config options, and Logfire query examples.