Anthropic’s agent benchmark shifts focus to end-to-end task success

Anthropic introduced a benchmark that evaluates AI agents on multi-step, tool-using workflows, emphasizing full-task completion over single-turn accuracy. The key shift is measuring long-horizon reliability and real-world execution (e.g., tool/API and possible UI flows), which better maps to production agent behavior.