Who contributed to OpenAI on May 28, 2026?

4 developers shipped this update, including aibrahim-oai, ningyi-oai, canvrno-oai, and romainhuet.

What were the notable OpenAI updates?

[codex] Add friendly Python SDK sandbox presets, [codex] add compaction metadata to turn headers, and fix(tui): complete vim word-end and line-end behavior (#24380).

OpenAI

Codex, the SDKs, and the engine behind ChatGPT

Pick a date

Topics: AI / ML Full archive →

The Wire · Showcase

CODEX FIXES VIM EDITING AND SANDBOX CHAOS IN PYTHON SDK

By RepoJournal · Filed 06:03 UTC on May 28, 2026 · About OpenAI

4 people shipped this

aibrahim-oai @aibrahim-oai 1 cited

ningyi-oai @ningyi-oai 1 cited

canvrno-oai @canvrno-oai 1 cited

romainhuet @romainhuet 1 cited

Codex shipped a suite of fixes that finally align Python SDK sandbox selection with actual developer workflows, while the TUI gets its vim keybindings straightened out.

The Python SDK's sandbox configuration just got a serious UX upgrade [1]. Threads and turns were using completely different APIs to do the same thing, burying wire-protocol details into what should be straightforward preset selection. That's gone now: one keyword, one preset name, consistent across the stack. The openai-node README also caught up to reality with gpt-5.5 and gpt-realtime-2 examples [6], so docs won't lead new callers astray. On the TUI side, Vim mode is now actually Vim [3]. The `e` key no longer gets stuck at word boundaries, the `C` command finally works for changing to line end, and multiline prompts don't lose their newlines in surprising places. There's also a quality-of-life fix for slash commands [4]: draft your prompt first, move to the start, complete a slash command around it, and your text stays intact instead of vanishing. The real production fix landed in SQLx [5]. Codex runs thread and memory state through bundled SQLite, and under heavy concurrent load with multiple subagents, the old libsqlite3-sys was hitting WAL-reset corruption. The bump picks up a newer bundled SQLite that should close that gap. Compaction metadata also got wired into turn headers [2], adding request_kind and window_id fields so you can actually track what kind of work your context window is doing.

Action items

→ Update Python SDK calls to use named sandbox presets instead of SandboxMode/SandboxPolicy openai/codex [plan]
→ Verify SQLx bump stability in staging with heavy concurrent subagent load openai/codex [monitor]
→ Update documentation examples to reference gpt-5.5 and gpt-realtime-2 openai/openai-node [plan]

References

[1] [codex] Add friendly Python SDK sandbox presets ↗ openai/codex
[2] [codex] add compaction metadata to turn headers ↗ openai/codex
[3] fix(tui): complete vim word-end and line-end behavior (#24380) openai/codex
[4] fix: Preserve draft text when completing argument-taking slash commands ↗ openai/codex
[5] Bump SQLx to pick up newer bundled SQLite (#24728) openai/codex
[6] Update README models to gpt-5.5 and gpt-realtime-2 ↗ openai/openai-node

Quick answers

What shipped in OpenAI on May 28, 2026?: Codex shipped a suite of fixes that finally align Python SDK sandbox selection with actual developer workflows, while the TUI gets its vim keybindings straightened out. In total, 32 commits, 32 pull requests, and 2 releases landed.
Who contributed to OpenAI on May 28, 2026?: 4 developers shipped this update, including aibrahim-oai, ningyi-oai, canvrno-oai, and romainhuet.
What were the notable OpenAI updates?: [codex] Add friendly Python SDK sandbox presets, [codex] add compaction metadata to turn headers, and fix(tui): complete vim word-end and line-end behavior (#24380).

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.