Amazon Bedrock AgentCore Evaluations: Evaluate Any Agent Framework via OpenTelemetry

It's a new AWS service that lets you evaluate AI agents regardless of the framework they're built on, as long as they emit OpenTelemetry telemetry.

The service supports LangGraph, LlamaIndex, OpenAI Agents SDK, Google ADK, Claude Agent SDK, and Strands Agents, plus any OpenTelemetry-compatible framework.

It allows teams to evaluate agents with unified standards without changing infrastructure, reducing costs and accelerating AI adoption in enterprises.

Yes, as long as the system emits OpenTelemetry telemetry, it can connect to the service regardless of hosting location.