Who contributed to OpenAI on May 9, 2026?

3 developers shipped this update, including pakrym-oai, jiamingz42, and aibrahim-oai.

What were the notable OpenAI updates?

[codex] Remove legacy after tool use hooks, [codex] Delete function-style apply_patch, and Remove ToolName display helper.

OpenAI

Codex, the SDKs, and the engine behind ChatGPT

Pick a date

Topics: AI / ML Full archive →

The Wire · Showcase

CODEX KILLS LEGACY HOOKS AND SIMPLIFIES TOOL DISPATCH

By RepoJournal · Filed 06:02 UTC on May 9, 2026 · About OpenAI

3 people shipped this

pakrym-oai @pakrym-oai 3 cited

jiamingz42 @jiamingz42 1 cited

aibrahim-oai @aibrahim-oai 1 cited

Dead code came off the critical path overnight—Codex ripped out the unused AfterToolUse hook infrastructure that was still firing on every tool execution, then consolidated apply_patch onto a single freeform surface.

The AfterToolUse hook system [1] was a phantom menace: the registry never populated handlers for it, but the dispatch logic stayed wired through core tool execution anyway. That's gone now, along with the unused payload types. Simultaneously, apply_patch just lost its function-style JSON registration path [2]—models and tests can no longer invoke it the old way, leaving only the freeform implementation as truth. This consolidation matters because tool identity was bleeding into string comparisons; ToolName.display() has been removed [3] in favor of structural ordering that forces teams to compare tool identity properly, not by rendered spelling. The desktop app-server can now request and attach DeviceCheck attestation tokens [4] for scoped ChatGPT Codex requests. One note: the Python runtime wheel publishing got reverted [5], so don't expect those binaries from the latest release tag.

Action items

→ Audit custom AfterToolUse hook implementations—they're gone from core dispatch openai/codex [plan]
→ Migrate any function-style apply_patch calls to freeform tool registration openai/codex [plan]
→ Review ToolName comparisons—string-based logic is now exposed openai/codex [monitor]

References

[1] [codex] Remove legacy after tool use hooks ↗ openai/codex
[2] [codex] Delete function-style apply_patch ↗ openai/codex
[3] Remove ToolName display helper ↗ openai/codex
[4] [codex] request desktop attestation from app ↗ openai/codex
[5] Revert "Publish Python runtime wheels on release" ↗ openai/codex

Quick answers

What shipped in OpenAI on May 9, 2026?: Dead code came off the critical path overnight—Codex ripped out the unused AfterToolUse hook infrastructure that was still firing on every tool execution, then consolidated apply_patch onto a single freeform surface. In total, 46 commits, 48 pull requests, and 6 releases landed.
Who contributed to OpenAI on May 9, 2026?: 3 developers shipped this update, including pakrym-oai, jiamingz42, and aibrahim-oai.
What were the notable OpenAI updates?: [codex] Remove legacy after tool use hooks, [codex] Delete function-style apply_patch, and Remove ToolName display helper.

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.