RepoJournal
OpenAI OpenAI
150 wires and counting

$ follow OpenAI

Keep up with OpenAI in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-06-01
stories 17
RSS Archive Reviews

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

CODEX FIXES SUBAGENT DATA MODEL, SHIPS CLOUD CONFIG LAYER STACK

By RepoJournal · Filed · About OpenAI · Composed from the cited sources · methodology

joeflorencio-openai owenlin0
2 people shipped this
joeflorencio-openai @joeflorencio-openai 3 cited owenlin0 @owenlin0 1 cited

Codex killed a critical data modeling bug where parent thread relationships got tangled with fork history, then landed the first three PRs of a five-part cloud-managed config overhaul.

The headline fix [1] stops overloading `forked_from_id` to mean two things at once. Guardian and review subagents can now track their parent thread independently from whether they forked the main thread's history. This lands a new `parent_thread_id` field on `SessionMeta` and exposes it through app-server proto, fixing a data model that was breaking subagent relationships downstream.

Meanwhile, the config work ships hard. PR 1 [2] landed the transport layer and bundle endpoint that replaces legacy cloud requirements. PR 2 [3] added the requirements layer composition engine with merge semantics and provenance tracking. PR 3 [4] made enterprise-managed config a first-class layer source with backend-supplied metadata preserved through diagnostics and hook attribution. Together, these three PRs unlock cloud-delivered config as a real system, not a side channel.

On the lighter side, plugin install got faster [5]. Codex stopped forcing directory refresh during auth checks and now uses normal caching instead, cutting unnecessary requests while keeping auth state current. Regression tests lock in the behavior.

$ ls openai/reviews/ # the step-back reads

the week in review Codex hands model defaults to the server the month in review Agents API lands in both OpenAI SDKs
Full archive → Weekly reviews →

Quick answers

What shipped in OpenAI on June 1, 2026?
Codex killed a critical data modeling bug where parent thread relationships got tangled with fork history, then landed the first three PRs of a five-part cloud-managed config overhaul. In total, 8 commits, 8 pull requests, and 1 releases landed.
Who contributed to OpenAI on June 1, 2026?
2 developers shipped this update, including owenlin0 and joeflorencio-openai.
What were the notable OpenAI updates?
store and expose parent_thread_id on Threads, Add config bundle transport types, and Compose requirements layers.
♥

$ log OpenAI --recent

Safety identifiers reach images, plugin approvals bind to contents 93 shipped Credential masking lands in Codex, Live gains Decisions across both SDKs 100 shipped Codex stabilizes environment tools, python adds residency options 135 shipped openai-python restores streaming connection reuse, browser and computer toolsets land in both SDKs 123 shipped

$ log --all # elsewhere on the wire

Alignment & Post-training verl rejects conflicting vLLM engine_kwargs, open-instruct adds MILES OPD behind a Megatron learner Interpretability nnsight 0.8.0 cuts import time from 7.6s to 1.6s

$ ls showcases/

Safety Evals AI Red Teaming Agent Security Interpretability Alignment & Post-training

$ init your-desk

RepoJournal can write one of these for any repo you care about. Sign in with GitHub, pick your repos and sources, and your desk gets a wire like this one every morning.