What shipped in OpenAI on May 17, 2026?

Python SDK now handles login natively, ending the workaround era for app authentication. In total, 10 commits and 10 pull requests landed.

Who contributed to OpenAI on May 17, 2026?

5 developers shipped this update, including aibrahim-oai, etraut-openai, sayan-oai, miz-openai, and bolinfest.

What were the notable OpenAI updates?

sdk/python: add first-class login support, [1 of 4] tui: route primary settings writes through app server, and multiagent: trim model-visible description, cap to 5 models.

OpenAI

Codex, the SDKs, and the engine behind ChatGPT

Pick a date

Topics: AI / ML Full archive →

The Wire · Showcase

CODEX SHIPS AUTH-FIRST SDK, LOCKS DOWN REMOTE CONFIG

By RepoJournal · Filed 06:02 UTC on May 17, 2026 · About OpenAI

5 people shipped this

aibrahim-oai @aibrahim-oai 1 cited

etraut-openai @etraut-openai 1 cited

sayan-oai @sayan-oai 1 cited

miz-openai @miz-openai 1 cited

bolinfest @bolinfest 1 cited

Python SDK now handles login natively, ending the workaround era for app authentication.

The Python SDK gained first-class login support [1], moving authentication from an external headache into a normal workflow. API-key flows complete immediately while interactive ChatGPT logins now route through the SDK, finally giving developers a clean setup path. In parallel, the TUI is being hardened against state drift: primary settings writes now route through the app server instead of writing directly to local config [2], the first of four planned migrations that will lock all preference changes into the centralized backend. Remote executor registration also switched to auth-backed credentials [4], replacing the old registry bearer env var with persisted ChatGPT auth, which streamlines credential management without breaking the existing biscuit-only contract. On the model side, the spawn_agent guidance got aggressive trimming [3]: model override summaries are now capped to 5 entries with compact reasoning, cutting unnecessary context bloat. Tests across the suite also cleaned up legacy compatibility shapes [5], replacing direct SandboxPolicy construction with canonical PermissionProfile builders to clarify where old patterns are actually required.

Action items

→ Integrate new Python SDK login flows into app setup paths openai/codex [plan]
→ Audit TUI remote session preference handling for upcoming config migrations openai/codex [monitor]
→ Update custom executor registration to use new auth path before next deploy openai/codex [plan]

References

[1] sdk/python: add first-class login support ↗ openai/codex
[2] [1 of 4] tui: route primary settings writes through app server ↗ openai/codex
[3] multiagent: trim model-visible description, cap to 5 models ↗ openai/codex
[4] exec-server: support auth-backed remote executor registration ↗ openai/codex
[5] test: construct permission profiles directly ↗ openai/codex

Quick answers

What shipped in OpenAI on May 17, 2026?: Python SDK now handles login natively, ending the workaround era for app authentication. In total, 10 commits and 10 pull requests landed.
Who contributed to OpenAI on May 17, 2026?: 5 developers shipped this update, including aibrahim-oai, etraut-openai, sayan-oai, miz-openai, and bolinfest.
What were the notable OpenAI updates?: sdk/python: add first-class login support, [1 of 4] tui: route primary settings writes through app server, and multiagent: trim model-visible description, cap to 5 models.

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.