Highlight
Featured this release
Test voice and dictation flows with microphone audio input
New Audio Input configs feed prepared recordings into your app's microphone mid-test, so agents can exercise dictation and voice-driven flows.
Voice interfaces used to be untestable end to end: an automated browser hands your app silence, so speech-to-text never fires.
Add an Audio Input config under Settings → Configs, upload a short recording and attach it to a test case. When the agent starts recording or opens the microphone, it plays that clip in, so speech-to-text and voice-command flows get real audio. The played clip appears as a playable track on that step in the run tracer.