Skip to content

fix(voice): redact STT prompts from traces - #4663

Merged
seratch merged 1 commit into
openai:mainfrom
hsusul:fix/voice-stt-prompt-trace-redaction
Aug 26, 2026
Merged

seratch merged 1 commit into
openai:mainfrom
hsusul:fix/voice-stt-prompt-trace-redaction

Conversation

@hsusul

@hsusul hsusul commented Aug 25, 2026

Copy link
Copy Markdown
Contributor

Summary

This pull request fixes OpenAI STT prompts appearing in trace metadata when trace_include_sensitive_data is disabled. Both static and streamed transcription paths now redact the prompt from model_config while continuing to send the original prompt to the provider.

Test plan

  • Added static transcription coverage proving the provider receives the prompt while the trace does not.
  • Extended streamed transcription coverage to verify prompt and keyword redaction together.
  • Ran .agents/skills/code-change-verification/scripts/run.sh; formatting, lint, type checking, and the full test suite passed.

Issue number

None.

Checks

  • I've added new tests, if relevant
  • I've run .agents/skills/code-change-verification/scripts/run.sh
  • I've confirmed all verification steps pass
  • If using Codex, I've run /review before submitting this PR

@seratch seratch added this to the 0.22.x milestone Aug 26, 2026
@seratch
seratch merged commit 1df6e81 into openai:main Aug 26, 2026
17 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants