openai / openai/codex

Voice dictation input is discarded when Codex asks a clarification question

Open
#38,195 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

app bug plan windows-os
Dominant language
Rust
Stars
125k
Forks
19.5k
PR merge metrics
PR metrics pending

Description

What version of the Codex App are you using (From “About Codex” dialog)?

Version 26.803.81509

What subscription do you have?

Business

What platform is your computer?

Microsoft Windows NT 10.0.26200.0 x64

What issue are you seeing?

When using voice dictation in the Codex Windows desktop app, an ongoing or recently completed voice input can be discarded if Codex displays a clarification question, for example while working in Plan Mode.

This is particularly frustrating with longer dictated prompts because the user may already have spoken for several seconds or longer before the clarification dialog appears. The dictated input is then lost and has to be recorded again.

What steps can reproduce the bug?
  1. Open the Codex desktop app on Windows.
  2. Start a task in Plan Mode where Codex may ask clarification questions.
  3. While Codex is working, start dictating a new message using voice input.
  4. Continue speaking while Codex is processing.
  5. Codex displays a clarification question.
  6. Observe that the current dictated input is discarded or no longer available.
What is the expected behavior?

Voice input that has already been recorded or transcribed should never be silently discarded when Codex displays a clarification question.

Additional information

Ideally, Codex should:

  • preserve the current dictation/transcription as a draft;
  • display the clarification question without replacing the user's input;
  • restore the draft after the clarification has been answered.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

No file or test is named. Start by reproducing the issue in the Windows desktop app through the voice-input and Plan Mode clarification flow, then trace how the dictated text is handled when the question appears. Done means recorded or transcribed input remains available as a draft and is restored after the clarification is answered.

Written by the indexing model from the issue text.

Assessment

Tech stack
rust
Domain
desktop
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
55/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.