Codex CLI 0.129.0 release
TL;DR
Codex CLI 0.129.0 adds modal Vim editing in the composer, redesigned TUI workflows for easier resuming and forking, enhanced plugin management with workspace sharing, and improved Linux sandbox reliability.
What changed
Codex CLI 0.129.0 landed with modal Vim editing inside the composer. Users can now switch into a full Vim mode without leaving the main input area.
The TUI received a redesign focused on resuming and forking sessions. Plugin management added workspace sharing so teams can push and pull plugin sets across machines.
Linux sandbox reliability improved through tighter file and process controls. These updates arrived in the May 19, 2026 release.
Why it matters
Solo builders who rely on Codex for daily coding now get fewer context switches when editing prompts or code snippets. The session resume and fork changes reduce the friction of starting from a previous state or branching an experiment.
Workspace plugin sharing lowers the setup cost for consistent environments across projects. Stronger Linux sandboxing addresses a common pain point for users running agents on remote servers or containers.
How to use it
Run npm install -g @openai/codex@0.129.0 to pull the update. Open any Codex session and press the new keybinding to enter modal Vim editing in the composer.
Check the updated TUI help menu for the redesigned resume and fork commands. Configure shared plugin workspaces through the new plugin management flags documented in the CLI reference.
Watch for
Stable adoption will show up as fewer reported sandbox escapes on Linux and higher usage of the fork command in public workflows. The next signal to track is whether OpenAI extends the same modal editing and workspace sharing to the IDE extension.
Harsh’s take
Codex CLI 0.129.0 gives solo operators better local control over sessions and plugins, but it still ties heavy lifting to OpenAI's backend models and rate limits. The Vim mode and TUI tweaks reduce daily friction, yet they do not remove the need to manage API spend or prompt drift over long tasks.
The real trade-off is speed versus ownership. Workspace sharing helps consistency, but it also increases the surface for config drift if you do not pin versions. Linux sandbox gains are welcome, yet they only matter if you already run agents outside macOS or Windows containers.
Do this now: update to 0.129.0, test a forked session on a real project, and decide whether the TUI changes save enough time to justify keeping Codex in your primary loop versus falling back to raw Claude Code or Cursor.
by Harsh Desai
About OpenAI Codex
View the full OpenAI Codex page →All OpenAI Codex updatesGo deeper
More AI news
- Daily RoundupGLM 5.3 and Vercel agent tools land, Gemini SAT practice, plus fresh models on Hugging Face
Google and Vercel released new agent and model features while Hugging Face and Replicate added trending models; OpenAI tightened safety after agent issues.
- Daily RoundupVercel-Origin beta and GPT-5.6 Sol discount, plus agent tools for builders
Vercel rolled out Origin repository deployments and a month-long GPT-5.6 Sol discount on AI Gateway while Product Hunt featured new agent layers and Hugging Face highlighted an uncensored Qwen model, with industry moves around data use and automation exits.
- Weekly DigestHermes Agent v0.20 voice and A2A protocol, OpenClaw mobile updates, and multi-agent tools to test now
Hermes Agent rolled out v0.20 releases with voice streaming, A2A interoperability, desktop artifacts, and self-recovery tools while OpenClaw added Android switching, iOS caching, and plugin provenance checks across the week.