Who contributed to OpenAI on July 31, 2026?

8 developers shipped this update, including mandal-OAI, jif, Charlie Marsh, Hayden, stainless-app, stainless-app[bot], jbeckwith-oai, and tescherm-openai.

What were the notable OpenAI updates?

release: 7.2.0, v2.51.0, and Wait for memory consolidation shutdown before finishing jobs (#36140).

OpenAI

Codex, the SDKs, and the engine behind ChatGPT

Pick a date

Topics: AI / ML Full archive →

The Wire · Showcase

BOTH SDKS SHIP FAST TIER SUPPORT AS CODEX TIGHTENS CONCURRENCY CONTROLS

By RepoJournal · Filed 06:58 UTC on July 31, 2026 · About OpenAI

8 people shipped this

Hayden @HAYDEN-OAI 2 cited

mandal-OAI @mandal-OAI 1 cited

jif @jif-oai 1 cited

Charlie Marsh @charliemarsh-oai 1 cited

stainless-app @stainless-app 1 cited

stainless-app[bot] @stainless-app[bot] 1 cited

jbeckwith-oai @jbeckwith-oai 1 cited

tescherm-openai @tescherm-openai 1 cited

The Node and Python SDKs landed fast tier APIs in coordinated releases overnight, while Codex hardened its memory consolidation and metadata handling against race conditions.

Node 7.2.0 and Python 2.51.0 both shipped with fast tier feature support [1][2], completing API parity across official clients. The Python SDK also moved the needle on reliability: it now honors Retry-After delays up to two minutes instead of capping at 60 seconds [4], a fix that prevents unnecessary thrashing during backend congestion. Separately, the Codex pipeline absorbed five critical stability improvements across 50 commits and 48 PRs. A phase-two memory job race condition got fixed by ensuring consolidation-agent shutdown completes before job completion [3], and concurrent metadata requests for the same remote path are now coalesced to eliminate duplicate RPCs [5]. Skill injection metrics now label explicit invocations separately [6] to surface the distinction that matters for debugging. The Node team also streamlined CI by removing the Agents SDK downstream check and JSR publishing [7][8], clearing out maintenance debt. Python added comprehensive mTLS examples for API-key authentication using native SSLContext [9], addressing a gap for teams running mutual TLS.

Action items

→ Upgrade to Node SDK 7.2.0 and Python SDK 2.51.0 to access fast tier openai/openai-node [plan]
→ Review Python Retry-After behavior if you have retry-sensitive workloads openai/openai-python [monitor]
→ Watch Codex memory consolidation metrics for stability gains openai/codex [monitor]

References

[1] release: 7.2.0 ↗ openai/openai-node
[2] v2.51.0 ↗ openai/openai-python
[3] Wait for memory consolidation shutdown before finishing jobs (#36140) openai/codex
[4] fix(client): honor Retry-After delays up to two minutes ↗ openai/openai-python
[5] Coalesce concurrent remote metadata requests (#36184) openai/codex
[6] Label explicit skill injection metrics (#36191) openai/codex
[7] ci: remove Agents SDK downstream check (#2033) openai/openai-node
[8] chore: remove JSR publishing (#2037) openai/openai-node
[9] docs: add API-key mTLS HTTP client recipes ↗ openai/openai-python

Quick answers

What shipped in OpenAI on July 31, 2026?: The Node and Python SDKs landed fast tier APIs in coordinated releases overnight, while Codex hardened its memory consolidation and metadata handling against race conditions. In total, 50 commits, 48 pull requests, and 2 releases landed.
Who contributed to OpenAI on July 31, 2026?: 8 developers shipped this update, including mandal-OAI, jif, Charlie Marsh, Hayden, stainless-app, stainless-app[bot], jbeckwith-oai, and tescherm-openai.
What were the notable OpenAI updates?: release: 7.2.0, v2.51.0, and Wait for memory consolidation shutdown before finishing jobs (#36140).

CODEX HARDENS SKILL SELECTION AND SESSION LOOKUP, NODE SDK CLOSES LINTER GAPS

Codex decouples explicit skill selection into a dedicated crate while tightening TUI session validation, as the Node SDK completes its oxlint cleanup stack across five rule categories.

ai 118 shipped

OpenAI Aug 5

CODEX HARDENS SECURITY, SDKS MODERNIZE TOOLCHAINS

Codex now prompts users before trusting local directories, closing a prompt injection exposure, while both Node and Python SDKs shift to Oxlint and Oxfmt tooling.

ai 114 shipped

OpenAI Aug 4

GPT-5.5 LANDS ACROSS NODE AND PYTHON SDKS, CODEX HARDENS MCP CONFORMANCE

OpenAI shipped gpt-5.5 model support and tool metadata across both JavaScript and Python SDKs while Codex locked down a full MCP client conformance regression suite to prevent protocol slippage.

ai 115 shipped

OpenAI Aug 3

CODEX HARDENS LOGIN FLOW WITH ONBOARDING HINTS AND METADATA PROTECTION

Three critical fixes shipped across auth, data persistence, and usage tracking to prevent thread metadata loss and secure OAuth callback handling.

ai 6 shipped

Elsewhere on the wire

AI Agents Aug 6

LANGCHAIN HARDENS ANTHROPIC TOOL CALLING AND GIT CREDENTIALS AS OPENHANDS SHIPS 1.10.0

LangChain fixed a critical credential leak in evaluation runs while Anthropic tool schemas get proper fallback handling, as three major agent frameworks push releases across the ecosystem.

ai-agents 54 shipped

Local LLMs Aug 6

OLLAMA FIXES KV CACHE CORRUPTION, LLAMA.CPP HARDENS FILE SEARCH, VLLM KILLS MOE LEGACY CODE

Ollama patched a critical cache corruption bug that corrupts lazy snapshots during multi-request prefills, while llama.cpp hardened Windows junction traversal and vLLM removed 14 hardcoded MoE shards across quantization methods.