Who contributed to OpenAI on July 25, 2026?

2 developers shipped this update, including copyberry and Justin Beckwith.

What were the notable OpenAI updates?

Make the Apps recovery exposure test deterministic, Enable resource reads for explicit executor skills, and Refresh MCP runtimes across thread startup.

OpenAI

Codex, the SDKs, and the engine behind ChatGPT

Pick a date

Topics: AI / ML Full archive →

The Wire · Showcase

CODEX HARDENS MCP RUNTIME STABILITY ACROSS FIVE CRITICAL FIXES

By RepoJournal · Filed 06:58 UTC on July 25, 2026 · About OpenAI

2 people shipped this

copyberry @copyberry 5 cited

Justin Beckwith @jbeckwith-oai 2 cited

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.

Five commits landed in Codex overnight, all targeting the same underlying problem: MCP configuration changes were racing against thread initialization and runtime refresh cycles [1][3][5]. The most urgent fix centralizes MCP runtime invalidation in ThreadManager so it catches threads mid-startup, preventing stale runtimes from persisting after config reload [3]. Companion fixes address authority staleness in elicitation reviews [4] and ensure managed MCP requirements propagate into active threads [5]. Earlier, a test determinism fix for Apps recovery [1] eliminates a race in the retry flow that was making test results unreliable, and executor skills now properly expose resource-access metadata so explicitly selected skills can read their own resources [2]. On the Node side, the test matrix gate consolidates CI coverage [6], letting the main branch rule replace version-specific checks with one stable context. Python SDK raised its floor to 3.10 [7], dropping Python 3.9 support after upstream EOL and automating monthly version-drift reviews to catch lifecycle changes early.

Action items

→ Review Codex thread startup behavior in staging to validate MCP refresh changes before next deploy openai/codex [plan]
→ Update Python 3.9 CI and deployment gates; codify minimum 3.10 in docs openai/openai-python [plan]
→ Merge Node test matrix gate to unblock branch protection simplification openai/openai-node [monitor]

References

[1] Make the Apps recovery exposure test deterministic ↗ openai/codex
[2] Enable resource reads for explicit executor skills ↗ openai/codex
[3] Refresh MCP runtimes across thread startup ↗ openai/codex
[4] Use current MCP authority for elicitation reviews ↗ openai/codex
[5] Refresh managed MCP requirements for active threads ↗ openai/codex
[6] ci: add stable Node test matrix gate ↗ openai/openai-node
[7] feat: require Python 3.10 and automate version reviews ↗ openai/openai-python

Quick answers

What shipped in OpenAI on July 25, 2026?: 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. In total, 31 commits, 31 pull requests, and 4 releases landed.
Who contributed to OpenAI on July 25, 2026?: 2 developers shipped this update, including copyberry and Justin Beckwith.
What were the notable OpenAI updates?: Make the Apps recovery exposure test deterministic, Enable resource reads for explicit executor skills, and Refresh MCP runtimes across thread startup.

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

OpenAI 4 days ago

CODEX TIGHTENS MEMORY HANDLING AND AUDIO PRESERVATION ACROSS THREADS

Codex shipped five structural fixes overnight that close gaps in memory reconciliation, audio attachment handling, and Windows permission checks that could silently corrupt state in production.

ai 105 shipped 1-min read

Elsewhere on the wire

AI Agents about 5 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 5 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.