Who contributed to OpenAI on May 2, 2026?

3 developers shipped this update, including jif-oai, xl-openai, and xli-oai.

What were the notable OpenAI updates?

feat: export and replay effective config locks, feat: Track local paths for shared plugins, and Add remote plugin skill read API.

OpenAI

Codex, the SDKs, and the engine behind ChatGPT

Pick a date

Topics: AI / ML Full archive →

The Wire · Showcase

CODEX TIGHTENS PLUGIN TRACKING AND KILLS LEGACY HOOK CRUFT

By RepoJournal · Filed 06:02 UTC on May 2, 2026 · About OpenAI

3 people shipped this

jif-oai @jif-oai 1 cited

xl-openai @xl-openai 1 cited

xli-oai @xli-oai 1 cited

Codex is shipping five architectural improvements tonight that lock down plugin reproducibility, add remote skill preview, and finally deprecate the legacy notify system that's been dead weight since the lifecycle hook engine took over.

The biggest move: effective config lockfiles [1] now let you export a session's resolved configuration and replay it exactly, solving the reproducibility nightmare where layered configs, CLI overrides, and feature flags made every run a mystery. Paired with that, local plugin path tracking [2] now maps remote shares to their local copies, so plugin/share/list shows you both the remote URL and where it's actually living on disk. On the client side, the new remote plugin skill read API [3] pulls markdown directly from plugin-service, letting the desktop UI preview uninstalled plugin skills before you commit the bandwidth to download them. The runtime is also getting leaner: goal continuation suppression is gone [5], which means /goal now keeps the agent working until it actually solves the problem instead of bailing early when it feels stuck. Finally, the legacy notify config system is officially deprecated [4], with a startup notice that'll help you spot which old plugins are still using it before the next cleanup pass.

Action items

→ Export your effective config lockfile on next session and commit it for reproducibility openai/codex [plan]
→ Audit your plugin configs for legacy notify usage — deprecation notices ship tonight openai/codex [monitor]
→ Test /goal continuation on complex multi-step problems — behavior changed openai/codex [monitor]

References

[1] feat: export and replay effective config locks ↗ openai/codex
[2] feat: Track local paths for shared plugins ↗ openai/codex
[3] Add remote plugin skill read API ↗ openai/codex
[4] deprecate legacy notify (#20524) openai/codex
[5] Remove no-tool goal continuation suppression (#20523) openai/codex

Quick answers

What shipped in OpenAI on May 2, 2026?: Codex is shipping five architectural improvements tonight that lock down plugin reproducibility, add remote skill preview, and finally deprecate the legacy notify system that's been dead weight since the lifecycle hook engine took over. In total, 31 commits, 31 pull requests, and 1 releases landed.
Who contributed to OpenAI on May 2, 2026?: 3 developers shipped this update, including jif-oai, xl-openai, and xli-oai.
What were the notable OpenAI updates?: feat: export and replay effective config locks, feat: Track local paths for shared plugins, and Add remote plugin skill read API.

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.