mirror of
https://github.com/openai/codex.git
synced 2026-09-13 11:47:17 +00:00
For better caching performance all output items should be rendered in the order they were produced before all new input items (for example, all function_call before all function_call_output).