gpt-5.6-sol incorrectly flags harmless prompts as policy violations #46093

Activity

1. zhoulingluo

   added

   bugSomething isn't workingSomething isn't working

   on Sep 17, 2026
2. github-actions

   added

   CLIIssues related to the Codex CLIIssues related to the Codex CLI

   model-behaviorIssues related to behaviors exhibited by the modelIssues related to behaviors exhibited by the model

   on Sep 17, 2026
3. github-actions commented on Sep 17, 2026

   

   github-actionsBot

   on Sep 17, 2026 – with GitHub Actions

   Contributor

   More actions

   Potential duplicates detected. Please review them and close your issue if it is a duplicate.

   • [Windows] GPT-6 Astra rejects every prompt with "Invalid prompt" #46082
   • gpt-6-astra rejects benign arithmetic in isolated CLI 0.154.0; gpt-5.6-sol succeeds #45993
   • Codex App falsely flags harmless prompts like "1+1" and "hello" as invalid prompts #45339
   • Codex App falsely flags harmless prompts like "1+1" and "hello" as invalid prompts #45336
   • Model restrictions #45963

   Powered by Codex Action
4. clchen3667 commented on Sep 17, 2026

   

   clchen3667

   on Sep 17, 2026 – with ChatGPT Codex Connector

   More actions

   Confirmed independently on macOS with codex-cli 0.154.0-alpha.6.2 and a ChatGPT Pro account.

   At 2026-09-17 03:45 UTC, this isolated request returned the same Invalid prompt: your prompt was flagged as potentially violating our usage policy error:

   codex exec --ephemeral --ignore-user-config --ignore-rules --skip-git-repo-check \
     -C /private/tmp -s read-only -m gpt-5.6-sol --json 'Reply only OK'

   This rules out workspace instructions, repository context, and the user config in our case. GPT-6 Astra had already been affected; GPT-5.6 Sol was initially usable and then began failing with the identical error later the same day.

   Local reproduction thread ID: 01a0ad77-ed62-7863-a9aa-d15a4ddf0a05.