Can I understand code by talking to it?

Yes. With DiffGuardian's Pro voice mode you ask a pull request or your whole codebase a question out loud and hear the answer. When you talk to a pull request, the code it explains is highlighted and scrolled into view. When you talk to your codebase, a graph builds on screen while the answer is spoken. It is the fastest way to a first understanding, with no reading required.

#Two ways to talk

• Talk to a pull request. Hold the key in the voice dock, ask what it does, and hear the walkthrough. Ask a follow-up, or skip to the file you care about. See Talk to your pull request.
• Talk to your codebase. In Analyze Project, ask how auth works or what a service touches, and watch the graph build as you listen. See Talk to your codebase.

#Why voice is quick

• You do not need to know where to look. Ask the question you actually have.
• You listen while you look. The highlighted code or the graph is on screen as you hear about it.
• Your hands stay free for the part that needs them.

#Set it up

Choose System, OpenAI or ElevenLabs in Settings → Voice, grant microphone access and hold the key shown in the dock. Details are in Voice.

DiffGuardian is a desktop app for macOS, Windows and Linux, and works with pull requests on GitHub, GitLab and Bitbucket.

#Limits

• Voice is a Pro feature. OpenAI and ElevenLabs use your own key. The system voice needs no key.
• Typing always works. Voice is the fast path, not the only one.
• It is separate from Trace It, which is the written walkthrough you read at your own pace.