Use Cases › Voice-Driven Code Review with OpenCode

🔍

Voice-Driven Code Review with OpenCode

The Problem

Code review is a conversation, but typing review comments breaks your reading flow. You switch between reading code and typing feedback.

The Solution

Use walkie-talkie mode to speak review comments while examining code. The AI processes your spoken feedback and responds with analysis.

Workflow

  1. Open a session and load the code you want reviewed in context.
  2. Read through the code while the AI is idle.
  3. Type /ptt and speak your observation: "This function has O(n²) complexity. Could we make it O(n log n)?"
  4. Type /ptt again. The AI receives your spoken comment and responds with analysis.
  5. Hear the response via TTS while continuing to look at the code.
  6. Interrupt with follow-up thoughts as they come: "What about the edge case when the input is null?"
Tips
• Walkie-talkie mode is perfect for code review because it matches the natural back-and-forth rhythm.
• Whisper handles technical terms like "O(n log n)", "recursive", "memoization" well with the small or medium model.
• Voice interruption lets you jump in when the AI misses the point — no need to wait for it to finish.

Want to try this workflow?

Install interrupt →