openai / openai/codex

Voice walkthrough presents a nonexistent form question as verified project state

Open
#43,641 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

app bug model-behavior
Dominant language
Rust
Stars
125k
Forks
19.4k
PR merge metrics
PR metrics pending

Description

Environment

Codex desktop app on Windows, active voice session, September 7, 2026. App version and voice-model identifier were not captured.

Incident

During a website-funnel review, the user approved the contact-permission step and asked to continue. The assistant said the next screen asked about tax liens and UCC filings, supplied answer choices, and requested approval. Checking the actual source confirmed that no such question existed; the next screen was confirmation. The assistant retracted the claim.

Expected behavior

Descriptions of existing screens should be grounded in the actual page or verified project source. Suggestions should be identified as suggestions.

Impact

The error disrupted the review and could cause unintended changes based on fabricated requirements.

Requested review

Investigate grounding during voice walkthroughs while another task performs project work. This is one observed incident; the internal cause and deterministic reproduction are unknown.

Privacy

This report excludes business records, project names, private transcript history, screenshots, and session identifiers.

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 files, tests, entry points, or deterministic reproduction are named. Start by locating the voice walkthrough and project-grounding paths, then review the incident details and existing tests; done means establishing a reproducible case and a verified behavior or documenting why the cause cannot yet be isolated.

Written by the indexing model from the issue text.

Assessment

Tech stack
rust
Domain
ai-infra-agents
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.