@mastra/arthur

@mastra/arthur - OpenTelemetry + OpenInference Tracing Exporter

Export Mastra traces to Arthur AI using OpenInference Semantic Conventions.

Installation

npm install @mastra/arthur

Configuration

Add ArthurExporter to your Mastra configuration to export traces to Arthur. The exporter automatically reads credentials from environment variables, enabling zero-config setup.

Zero-Config

Set environment variables:

ARTHUR_API_KEY=your-api-key
ARTHUR_BASE_URL=https://app.arthur.ai
ARTHUR_TASK_ID=your-task-id  # optional, associates traces with a specific Arthur task

import { ArthurExporter } from '@mastra/arthur';
import { Mastra } from '@mastra/core/mastra';

const mastra = new Mastra({
  ...,
  observability: {
    configs: {
      arthur: {
        serviceName: 'my-service',
        exporters: [new ArthurExporter()],
      },
    },
  },
});

Explicit Configuration

import { ArthurExporter } from '@mastra/arthur';
import { Mastra } from '@mastra/core/mastra';

const mastra = new Mastra({
  ...,
  observability: {
    configs: {
      arthur: {
        serviceName: 'my-service',
        exporters: [
          new ArthurExporter({
            apiKey: 'your-api-key',
            endpoint: 'https://app.arthur.ai',
          }),
        ],
      },
    },
  },
});

Optional Configuration

new ArthurExporter({
  // Required at runtime (or set ARTHUR_API_KEY env var)
  apiKey: 'your-api-key',
  // Required at runtime (or set ARTHUR_BASE_URL env var)
  endpoint: 'https://app.arthur.ai',
  // Optional headers added to each OTLP request
  headers: {
    'x-custom-header': 'value',
  },
  // Optional log level for debugging
  logLevel: 'debug',
  // Optional batch size for the underlying BatchSpanProcessor
  batchSize: 512,
  // Optional timeout for span export
  timeout: 30000,
  // Optional resource attributes added to each span
  resourceAttributes: {
    'custom.attribute': 'value',
  },
});

Custom metadata

Custom span attributes are serialized into the OpenInference metadata payload. Add them through tracingOptions.metadata:

await agent.generate(input, {
  tracingOptions: {
    metadata: {
      companyId: 'acme-co',
    },
  },
});

OpenInference Semantic Conventions

This exporter follows the OpenInference Semantic Conventions for generative AI applications. All agent runs, tool calls, and LLM generations are automatically tagged with the correct span kinds and attributes.

License

Apache 2.0