Introducing: Testing voice agents like you test your chat agents

LangWatch Scenario now offers native support for testing voice agents using the same methodology applied to text agents, including simulated callers, traces, and judge-based evaluation. The platform enables spec-driven voice agent development across multiple provider stacks, allowing teams to run executable scenarios that verify agent behavior rather than relying on manual testing. This capability proves especially valuable for coding agents that need reliable test harnesses to iteratively improve voice agent performance.