[AI-SDK] How to trace runs from new Agent class of v5 (and ToolLoopAgent in v6)

[arthberman]

wrapAISDK seems to not support new Agent class from v5 of ai-sdk (that will be ToolLoopAgent class in v6)
Any guidance on how to trace it properly with Braintrust? I currently wrap my calls with traced() to group them in the same trace, so I don't know how the OTEL approach can help me as I need to group LLM runs within the same trace

Example of new Agent class usage:

import { Experimental_Agent as Agent, stepCountIs, tool } from 'ai';
import { z } from 'zod';

const weatherAgent = new Agent({
  model: 'openai/gpt-4o',
  tools: {
    weather: tool({
      description: 'Get the weather in a location (in Fahrenheit)',
      inputSchema: z.object({
        location: z.string().describe('The location to get the weather for'),
      }),
      execute: async ({ location }) => ({
        location,
        temperature: 72 + Math.floor(Math.random() * 21) - 10,
      }),
    }),
    convertFahrenheitToCelsius: tool({
      description: 'Convert temperature from Fahrenheit to Celsius',
      inputSchema: z.object({
        temperature: z.number().describe('Temperature in Fahrenheit'),
      }),
      execute: async ({ temperature }) => {
        const celsius = Math.round((temperature - 32) * (5 / 9));
        return { celsius };
      },
    }),
  },
  stopWhen: stepCountIs(20),
});

const result = await weatherAgent.generate({
  prompt: 'What is the weather in San Francisco in celsius?',
});

console.log(result.text); // agent's final answer
console.log(result.steps); // steps taken by the agent