What shipped in OpenAI on May 1, 2026?

OpenAI

Codex, the SDKs, and the engine behind ChatGPT

Pick a date

The Wire · Showcase

CODEX CLOSES WINDOWS SANDBOX VULNERABILITY AS PLUGIN SYNC GOES LIVE

By RepoJournal · Filed 15:25 UTC on May 1, 2026 · About OpenAI

A critical Windows named-pipe ACL bypass in the elevated sandbox is patched [ref:1], while remote plugin bundle synchronization ships to keep installations in sync across backends [ref:2].

The Windows sandbox fix [1] closes a privilege escalation window where elevated Codex processes could access named pipes created by build tools like Ninja without proper restrictions. The patch adds elevated-only token constructors that lock down the restricted SID list — a sharp security move that leaves the unelevated path untouched, minimizing blast radius. Meanwhile, remote plugin bundle syncing [2] now downloads and upgrades cached plugin bundles during app-server startup, keeping local installations aligned with backend versions without polluting config.toml. On the analytics front [4], remote plugin installs now emit proper telemetry that preserves marketplace identity while correctly tagging backend-sourced plugins, giving ops real visibility into plugin adoption patterns. Vim mode lands in the TUI composer [5], letting modal editors stop fighting Codex keybindings and stay in muscle memory during prompt drafting. One note: the thread/turns/list API [3] is deliberately marked experimental and pulled from public consumption — there are bugs to work out.

Action items

→ Deploy Windows sandbox patch [ref:1] to all elevated environments immediately openai/codex [immediate]
→ Test remote plugin bundle sync [ref:2] against your backend before general rollout openai/codex [plan]
→ Monitor remote plugin install analytics [ref:4] for correct plugin_id override after deploy openai/codex [monitor]
→ Do not consume thread/turns/list API [ref:3] — it's experimental and unstable openai/codex [immediate]

References

[1] Windows sandbox named-pipe ACL fix
[2] Remote plugin bundle sync
[3] Thread/turns/list API marked experimental
[4] Remote plugin install analytics
[5] Vim mode in TUI composer

Quick answers

What shipped in OpenAI on May 1, 2026?: A critical Windows named-pipe ACL bypass in the elevated sandbox is patched [ref:1], while remote plugin bundle synchronization ships to keep installations in sync across backends [ref:2]. In total, 45 commits, 45 pull requests, and 3 releases landed.

CODEX HARDENS MCP RUNTIME STABILITY ACROSS FIVE CRITICAL FIXES

OpenAI shipped a coordinated hardening of the Model Context Protocol layer across Codex, patching race conditions in thread startup, resource access, and authority management that could leave stale configurations in production.

ai 66 shipped 1-min read

OpenAI 1 day ago

SDK RELEASES LOCK IN PROMPT CACHING AND SPEND CONTROLS

Both Node and Python clients shipped v6.49.0 and v2.48.0 overnight with admin spend limit APIs and relaxed prompt cache validation.

ai 103 shipped 1-min read

OpenAI 2 days ago

CODEX CONSOLIDATES THREAD STARTUP, NODE SDK ADDS REALTIME AND STANDARD SCHEMA SUPPORT

Codex refactored its threading and database layer for consistency while the Node SDK ships four new schema helpers and fixes audio streaming in a single release push.

ai 120 shipped 1-min read

OpenAI 3 days ago

CODEX FIXES WINDOWS EXECUTION BUGS WHILE TIGHTENING SECURITY REVIEW GATES

Codex shipped critical fixes for Windows process management and terminal navigation that were silently breaking execution sessions on Windows machines.

ai 104 shipped 1-min read

Elsewhere on the wire

AI Agents about 9 hours ago

CLAUDE OPUS 5 LANDS ACROSS THE STACK

The newest Anthropic model is now live in langchain, Cline, and llama-index, with native support for extended reasoning and 1M context windows.

ai-agents 28 shipped 1-min read

Local LLMs about 9 hours ago

OLLAMA LANDS LAGUNA SUPPORT AND CRUSHES MEMORY LEAKS WHILE SGLANG HITS V0.5.16 WITH CONFIDENCE-DRIVEN SPECULATIVE DECODING

Ollama shipped three critical performance and reliability fixes for Metal residency and concurrent access patterns, while SGL-Lang released 0.5.16 with a new speculative algorithm hitting 383.7 tok/s on DeepSeek-V4.