Key Info
OpenAI's Tibo posted Day 19 small updates for Codex, covering faster call connections, cache behavior when changing reasoning mid-conversation, and a batch of Codex CLI improvements.
Highlights
- Calls connect much faster for dots.
- Changing reasoning in Codex and Work mid-conversation no longer invalidates the cache.
- Codex CLI adds reusable prompt presets, inline model and permission settings, and enterprise MCP sign-in.
- Fixed failures when resuming long sessions and added Windows corporate-proxy authentication.
- CLI usage has grown 2X since DevDay.