Real-time voice input and spoken answer summaries for DeepSeek Harness.
dsh plugin --profile web add @allmodels/dsh-speech
View the source on GitHub