GenAI catch-up 調査ノート 2026-10-01

🤖 Generated by Claude Opus 5.5

候補と選定

1. 🤖 OpenAI が GPT-6 Astra で動く常時稼働エージェント dots を出し、専用のクラウドコンピューター、読み取り専用の自律調査、自動レビュー付きの操作を備え、連続タスクでスコープ逸脱が増えることも測定した

収集した情報

[A] OpenAI — Introducing dots (2026-09-29)

https://openai.com/index/introducing-dots/

[A] OpenAI — DevDay 2026 Recap (2026-09-29)

https://openai.com/index/devday-2026-recap/

[A] OpenAI Deployment Safety Hub — GPT-6 Astra System Card, Appendix: dots (2026-09-29)

https://deploymentsafety.openai.com/gpt-6-astra/sec:appendix-dots

[A] OpenAI Help Center — Getting started with your dot (2026-09-29、検索結果)

https://help.openai.com/en/articles/20001530-getting-started-with-your-dot

[B] ITmedia AI+ — OpenAI、「24時間働く」エージェント「dots」発表 Astra搭載、クラウド上で常時稼働 (2026-09-30)

https://www.itmedia.co.jp/aiplus/article/2609/30/2000001864/

[B] Simon Willison — OpenAI DevDay 2026 live blog (2026-09-29)

https://simonwillison.net/2026/Sep/29/openai-devday-2026-live-blog/

[B] Latent Space (AINews) — OpenAI DevDay 2026: Dots, 6.1 Sol, Ultrafast, Decisions API, Agents API, Spaces, Marketplace, and 1.2 Billion ChatGPT WAU (2026-09-30)

https://www.latent.space/p/ainews-openai-devday-2026-dots-61

[B] Latent Space (podcast) — Why Dwarkesh is Wrong about Computer Use + How OpenAI shipped its Jev competitor in 1 Week (2026-09-30)

https://www.latent.space/p/devday-2026

[B] VentureBeat — OpenAI launches Dots, always-on AI agent coworkers, and ChatGPT Space where they can collaborate with human teams (2026-09-29)

https://venturebeat.com/technology/openai-launches-dots-always-on-ai-agent-coworkers-and-chatgpt-space-where-they-can-collaborate-with-human-teams

[B] TechCrunch — OpenAI launches Dots, its bubbly agentic avatar (2026-09-29)

https://techcrunch.com/2026/09/29/openai-launches-dots-its-bubbly-agentic-avatar/

[B] NBC News — OpenAI launches Dots AI agents amid safety questions (2026-09-29)

https://www.nbcnews.com/tech/tech-news/openai-launches-dots-ai-agents-safety-questions-rcna600338

[B] SiliconANGLE — OpenAI launches Dots, always-on AI agents in ChatGPT with their own cloud computers (2026-09-29)

https://siliconangle.com/2026/09/29/openai-launches-dots-always-on-ai-agents-in-chatgpt-with-their-own-cloud-computers/

[B] BankInfoSecurity — OpenAI Dots Pushes Always-on Agents Into the Enterprise (2026-09-29)

https://www.bankinfosecurity.com/openai-dots-pushes-always-on-agents-into-enterprise-a-32970

[C] Flavio Copes — A deep dive into OpenAI dots (2026-09-30)

https://flaviocopes.com/openai-dots/

[C] DataCamp — OpenAI Dots: Always-On Agents in ChatGPT, Explained (2026-09-29)

https://www.datacamp.com/blog/openai-dots

[C] Hacker News — Dots: Always-on agents (2026-09-29、750 ポイント、629 コメント)

https://news.ycombinator.com/item?id=49896604

メモ

実践やアウトプットの提案

2. 🖥️ OpenAI が Codex のハーネスをマネージドで提供する Agents API にコンピューター操作を加え、AWS はツールを顧客側の計算資源で動かす AWS ネイティブ版の Bedrock Managed Agents をプレビューで公開した

収集した情報

[A] OpenAI — DevDay 2026 Recap (2026-09-29)

https://openai.com/index/devday-2026-recap/

[A] OpenAI — Introducing the Agents API (2026-09-10、背景)

https://openai.com/index/introducing-the-agents-api/

[A] OpenAI Developers — Computer use in the Agents API (2026-10-01 取得時点)

https://developers.openai.com/api/docs/guides/agents-api/tools/computer-use

[A] OpenAI Developers — Agents API changelog, overview, architecture, environments and tools (2026-10-01 取得時点)

https://developers.openai.com/api/docs/guides/agents-api/overview

[A] OpenAI Developers — Bedrock Managed Agents (2026-10-01 取得時点)

https://developers.openai.com/api/docs/guides/agents-api/bedrock-managed-agents

[A] AWS — Amazon Bedrock Managed Agents, powered by OpenAI, is now available in preview (2026-09-29)

https://aws.amazon.com/about-aws/whats-new/2026/09/bedrock-managed-agents-preview/

[A] AWS Docs — Amazon Bedrock Managed Agents, powered by OpenAI (preview): overview, limitations, setup, security and API reference (2026-10-01 取得時点)

https://docs.aws.amazon.com/bedrock/latest/userguide/bedrock-managed-agents-openai.html

[A] About Amazon — OpenAI models on Amazon Bedrock (背景)

https://www.aboutamazon.com/news/aws/bedrock-openai-models

[B] Simon Willison — OpenAI DevDay 2026 live blog (2026-09-29)

https://simonwillison.net/2026/Sep/29/openai-devday-2026-live-blog/

[B] Latent Space (podcast) — Why Dwarkesh is Wrong about Computer Use + How OpenAI shipped its Jev competitor in 1 Week (2026-09-30)

https://www.latent.space/p/devday-2026

[B] The Next Web — 'Data never leaves AWS': OpenAI agents move into Amazon's cloud (2026-09-29)

https://thenextweb.com/news/openai-bedrock-managed-agents-aws-devday

[B] The Decoder — OpenAI expands Codex and its API at DevDay with security scans, a Decisions API, and Ultrafast (2026-09-29)

https://the-decoder.com/openai-expands-codex-and-its-api-at-devday-with-security-scans-a-decisions-api-and-ultrafast/

[B] Decrypt — OpenAI Gave AI Agents Their Own Computers at DevDay 2026 (2026-09-29)

https://decrypt.co/379584/openai-ai-agents-computers-devday-2026-everything-announced

[B] gihyo.jp — OpenAI、AIとチームで作業する「ChatGPT Space」と、継続的に動作するクラウドエージェント「dots」を発表 GPT-6.1 Solも提供開始 (2026-09-30)

https://gihyo.jp/article/2026/09/openai-devday-20206-with-gpt-6.1-sol-and-dots

[B] npaka (note) — OpenAI DevDay 2026 発表まとめ - 開発者向けに押さえたい主要アップデート (2026-09-30)

https://note.com/npaka/n/n718a0b7acf3b

[C] Build with AWS (Substack) — Bedrock Managed Agents, explained like you're five (2026-09-29)

https://buildwithaws.substack.com/p/bedrock-managed-agents-explained

[C] Zenn (suwa-sh) — アプリケーションからCodexハーネスへ仕事を渡すOpenAI Agents API (2026-09-12、背景)

https://zenn.dev/suwash/articles/openai-codex-agents-api-openai-p1_20260911

[C] Hacker News — OpenAI Agents API (2026-09-10、350 ポイント、187 コメント、背景)

https://news.ycombinator.com/item?id=49649213

メモ

実践やアウトプットの提案

3. ♊ Google が出力上限 100 万トークンの Gemini 4 Argon を発表し、自社の比較表で最多の首位を示したが、提供はまず審査を経たサイバー防御組織に限った

収集した情報

[A] Google (The Keyword) — Gemini 4 Argon: our next era of frontier intelligence, by Koray Kavukcuoglu (2026-09-30)

https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/

[A] Google DeepMind — Gemini models page, Gemini 4 Argon comparison table (2026-09-30)

https://deepmind.google/models/gemini/

[A] Google DeepMind — Gemini 4 Argon with cybersecurity defense capabilities, and the Fairwind Program (2026-09-30)

https://deepmind.google/models/gemini/cyber/

[A] Google AI for Developers — Gemini API models, pricing and changelog (2026-10-01 取得時点)

https://ai.google.dev/gemini-api/docs/models

[B] VentureBeat — Google unveils Gemini 4 Argon, retaking benchmark lead over OpenAI and Anthropic — but in limited release (2026-09-30)

https://venturebeat.com/technology/google-unveils-gemini-4-argon-retaking-benchmark-lead-over-openai-and-anthropic-but-in-limited-release

[B] TechCrunch — Google releases Gemini 4 Argon, called its most powerful model yet (2026-09-30)

https://techcrunch.com/2026/09/30/google-releases-gemini-4-argon-called-its-most-powerful-model-yet/

[B] The Decoder — Google Gemini 4 Argon closes the gap with OpenAI and Anthropic but doesn't take a clear lead (2026-10-01)

https://the-decoder.com/google-gemini-4-argon-closes-the-gap-with-openai-and-anthropic-but-doesnt-take-a-clear-lead/

[B] Artificial Analysis — Gemini 4 Argon: Google is back as one of the top three labs (2026-09-30)

https://artificialanalysis.ai/articles/gemini-4-argon-google-top-three-labs

[B] Techzine Global — Gemini 4 "Argon" unveiled, but not yet released: what's going on? (2026-09-30)

https://www.techzine.eu/news/applications/144683/gemini-4-argon-unveiled-but-not-yet-released-whats-going-on/

[B] 9to5Google — Google announces Gemini 4 Argon as its new frontier model (2026-09-30)

https://9to5google.com/2026/09/30/gemini-4-argon-announcement/

[B] ITmedia NEWS — Google、「Gemini 4 Pro」ではなく「Gemini 4 Agron」発表 まずはサイバー防御組織に限定提供、出力上限は100万トークンに (2026-10-01)

https://www.itmedia.co.jp/news/article/2610/01/2000001912/

[B] ケータイ Watch (Impress) — グーグル、新AI「Gemini 4 Argon」発表、出力100万トークンへ拡大 (2026-10-01)

https://k-tai.watch.impress.co.jp/docs/news/2144686.html

[C] Hacker News — Gemini 4 Argon (2026-09-30、958 ポイント、653 コメント)

https://news.ycombinator.com/item?id=49913571

メモ

実践やアウトプットの提案

4. ☀️ OpenAI が GPT-6 Astra に迫る性能を 5 分の 1 の単価で出す GPT-6.1 Sol を公開し、1 週間で GPT-6 Sol を置き換え、同じ日に Codex・Bedrock・Copilot にも届いた

前回からの継続

収集した情報

[A] OpenAI — Introducing GPT-6.1 Sol (2026-09-29)

https://openai.com/index/introducing-gpt-6-1-sol/

[A] OpenAI Deployment Safety Hub — Addendum to GPT-6 Astra System Card: GPT-6.1 Sol (2026-09-29)

https://deploymentsafety.openai.com/gpt-6-1-sol

[A] OpenAI API docs — GPT-6.1 Sol, GPT-6 Sol and GPT-6 Astra model pages (2026-10-01 取得時点)

https://developers.openai.com/api/docs/models/gpt-6.1-sol

[A] OpenAI — Codex CLI 0.159.1 release and the Codex changelog (2026-09-29)

https://github.com/openai/codex/releases/tag/rust-v0.159.1

[A] AWS — GPT-6.1 Sol on Amazon Bedrock: announcement, blog post and model card (2026-09-29)

https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-6-1-sol.html

[A] GitHub — GPT-6.1 Sol in GitHub Copilot (2026-09-29)

https://github.blog/changelog/2026-09-29-gpt-6-1-sol-in-github-copilot

[B] Artificial Analysis — GPT-6.1 Sol replaces GPT-6 Sol after just 7 days, with near-Astra intelligence (2026-09-29)

https://artificialanalysis.ai/articles/gpt-6-1-sol-replaces-gpt-6-sol-after-just-7-days-with-near-astra-intelligence

[B] Simon Willison — OpenAI DevDay 2026 live blog (2026-09-29)

https://simonwillison.net/2026/Sep/29/openai-devday-2026-live-blog/

[B] Simon Willison — Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price war (2026-09-22、背景)

https://simonwillison.net/2026/Sep/22/opus-and-sol-and-luna/

[B] Latent Space (AINews) — OpenAI DevDay 2026: Dots, 6.1 Sol, Ultrafast, Decisions API, Agents API, Spaces, Marketplace, and 1.2 Billion ChatGPT WAU (2026-09-30)

https://www.latent.space/p/ainews-openai-devday-2026-dots-61

[B] TechCrunch — OpenAI launches GPT-6.1 Sol, says it nearly matches GPT-6 Astra and costs less (2026-09-29)

https://techcrunch.com/2026/09/29/openai-launches-gpt-6-1-sol-says-it-nearly-matches-gpt-6-astra-and-costs-less/

[B] The Next Web — OpenAI releases GPT-6.1 Sol at a fifth of GPT-6 Astra's token prices (2026-09-29)

https://thenextweb.com/news/openai-gpt-6-1-sol-price-astra-devday

[B] SiliconANGLE — OpenAI's GPT-6.1 Sol delivers Astra-like performance at a dramatically lower price (2026-09-29)

https://siliconangle.com/2026/09/29/openais-gpt-6-1-sol-delivers-astra-like-performance-at-a-dramatically-lower-price/

[B] Impress AI Watch — OpenAI、Astra並み性能を5分の1の価格で実現した「GPT-6.1 Sol」をリリース (2026-09-30)

https://ai.watch.impress.co.jp/docs/news/2144361.html

[B] CNET Japan (via Yahoo! News Japan) — OpenAI、「GPT-6.1 Sol」発表 Astraに迫る性能を5分の1の料金で (2026-09-30)

https://news.yahoo.co.jp/articles/8c316272aef79c370bf0c399aeef590e42fba9a0

[B] Classmethod DevelopersIO — Amazon Bedrock の GPT-6.1 Sol を試してみた (2026-10-01)

https://dev.classmethod.jp/articles/bedrock-openai-gpt61-sol/

[C] Zenn (schroneko) — OpenAI DevDay 2026 発表まとめ (2026-09-30)

https://zenn.dev/schroneko/articles/openai-devday-2026

[C] Qiita (mamagotolab) — Codex の新モデル GPT-6.1 Sol を GPT-6 Sol と実測比較した──トークン内訳・Plus の使用枠・codex review の指摘 (2026-09-30)

https://qiita.com/mamagotolab/items/42057aba50936c100e62

[C] Hacker News — GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price (2026-09-29、1,050 ポイント、931 コメント)

https://news.ycombinator.com/item?id=49896586

[C] Hacker News — GPT-6.1 Sol replaces GPT-6 Sol after just 7 days (2026-09-30、79 ポイント、97 コメント)

https://news.ycombinator.com/item?id=49906669

メモ

実践やアウトプットの提案

5. 🔀 OpenAI が Jev への対抗として数週間で作った、GPT-6 Luna に固定の選択肢で答えさせる Decisions API をプレビューで示し、Every の初期テストでは精度は Jev と同等、速度は結果が分かれた

収集した情報

[A] OpenAI — DevDay 2026 Recap (2026-09-29)

https://openai.com/index/devday-2026-recap/

[A] OpenAI on X — the OpenAI Developers announcement and Thibault Sottiaux's post (2026-09-29、検索結果)

https://x.com/OpenAIDevs/status/2105003318917697873

[A] OpenAI Developers — API changelog and documentation index (2026-10-01 取得時点)

https://developers.openai.com/api/docs/changelog

[B] Latent Space (podcast) — Why Dwarkesh is Wrong about Computer Use + How OpenAI shipped its Jev competitor in 1 Week (2026-09-30)

https://www.latent.space/p/devday-2026

[B] Latent Space (AINews) — OpenAI DevDay 2026: Dots, 6.1 Sol, Ultrafast, Decisions API, Agents API, Spaces, Marketplace, and 1.2 Billion ChatGPT WAU (2026-09-30)

https://www.latent.space/p/ainews-openai-devday-2026-dots-61

[B] Simon Willison — OpenAI DevDay 2026 live blog (2026-09-29)

https://simonwillison.net/2026/Sep/29/openai-devday-2026-live-blog/

[B] Every — Vibe Check: OpenAI DevDay 2026, by Dan Shipper (2026-09-29)

https://every.to/vibe-check/vibe-check-openai-devday-2026

[B] The New Stack — OpenAI answers TypeSafe's Jev with a Decision API built on Luna, by Frederic Lardinois (2026-09-29、検索結果)

https://thenewstack.io/openai-decision-api-luna/

[B] TechCrunch — OpenAI's Jev clone could help the frontier lab stop its swarming agents, by Tim Fernholz (2026-09-30)

https://techcrunch.com/2026/09/30/openais-jev-clone-could-help-the-frontier-lab-stop-its-swarming-agents/

[B] The Decoder — OpenAI expands Codex and its API at DevDay with security scans, a Decisions API, and Ultrafast (2026-09-29)

https://the-decoder.com/openai-expands-codex-and-its-api-at-devday-with-security-scans-a-decisions-api-and-ultrafast/

[C] Vercel — What is OpenAI's Decisions API? (2026-09-30)

https://vercel.com/i/what-is-openai-decisions-api

[C] OrcaRouter — OpenAI's Decisions API: GPT-6 Luna Picks One Answer (2026-09-30)

https://www.orcarouter.ai/blog/openai-decisions-api-gpt-6-luna

[C] Hacker News — Decisions API threads (2026-09-29 〜 2026-09-30)

https://news.ycombinator.com/item?id=49897587

[C] GitHub — agentjido/req_llm issue #1062 (2026-09-29)

https://github.com/agentjido/req_llm/issues/1062

[C] AI Heartland — Decisions APIとは|Jevの対抗馬と同名OSS実装の差をスキーマ15ケースで実測 (2026-09-30)

https://ai-heartland.com/llm/decisions-api/

メモ

実践やアウトプットの提案

6. ☁️ OpenAI の DevDay での Codex の更新は、再利用できるクラウド環境、エージェント管理画面・音声・worktree を備えた CLI、AGENTS.md にルールを書くコードレビュー、Codex Security Cloud をそろえた

収集した情報

[A] OpenAI — DevDay 2026 Recap, Codex section (2026-09-29)

https://openai.com/index/devday-2026-recap/

[A] OpenAI — Codex Cloud overview and cloud environments (2026-10-01 取得時点)

https://learn.chatgpt.com/docs/environments/cloud-environments.md

[A] OpenAI — Codex CLI 0.159.0 release, CLI documentation and source at rust-v0.159.0 (2026-09-29)

https://github.com/openai/codex/releases/tag/rust-v0.159.0

[A] OpenAI — Code Review and repository review documentation (2026-10-01 取得時点)

https://learn.chatgpt.com/docs/third-party/github.md

[A] OpenAI — Codex Security documentation and the openai/codex-security repository (2026-10-01 取得時点)

https://learn.chatgpt.com/docs/security/setup.md

[A] OpenAI — ChatGPT and Codex changelog (2026-10-01 取得時点)

https://learn.chatgpt.com/docs/changelog

[B] Simon Willison — OpenAI DevDay 2026 live blog (2026-09-29)

https://simonwillison.net/2026/Sep/29/openai-devday-2026-live-blog/

[B] Latent Space (AINews) — OpenAI DevDay 2026: Dots, 6.1 Sol, Ultrafast, Decisions API, Agents API, Spaces, Marketplace, and 1.2 Billion ChatGPT WAU (2026-09-30)

https://www.latent.space/p/ainews-openai-devday-2026-dots-61

[B] TechCrunch — OpenAI gives Codex reusable cloud environments that work across devices, by Sarah Perez (2026-09-29)

https://techcrunch.com/2026/09/29/openai-gives-codex-reusable-cloud-environments-that-work-across-devices/

[B] SiliconANGLE — OpenAI's Codex gets reusable cloud environments that follow developers across devices (2026-09-29)

https://siliconangle.com/2026/09/29/openais-codex-gets-reusable-cloud-environments-that-follow-developers-across-devices/

[B] The Decoder — OpenAI expands Codex and its API at DevDay with security scans, a Decisions API, and Ultrafast (2026-09-29)

https://the-decoder.com/openai-expands-codex-and-its-api-at-devday-with-security-scans-a-decisions-api-and-ultrafast/

[B] Developer Tech — OpenAI Codex Security Cloud reviews new GitHub commits (2026-09-30)

https://www.developer-tech.com/news/openai-codex-security-cloud-reviews-new-github-commits/

[B] GIGAZINE — OpenAIが「Codex CLI」を大幅刷新&Codexをクラウド上で動作させる「Codex Cloud」も登場 (2026-09-30)

https://gigazine.net/news/20260930-openai-codex-cli-cloud/

[B] ITmedia NEWS — OpenAIの「DevDay 2026」で発表された主なことまとめ (2026-09-30)

https://www.itmedia.co.jp/news/article/2609/30/2000001868/

[C] Zenn (schroneko) — OpenAI DevDay 2026 発表まとめ (2026-09-30)

https://zenn.dev/schroneko/articles/openai-devday-2026

[C] Zenn (kuu) — OpenAI DevDay 2026 keynote まとめ (2026-09-30)

https://zenn.dev/sushi/articles/1afa7489714951

[C] Hacker News — DevDay 2026 Recap (2026-09-29、94 ポイント、48 コメント) と Codex 関連の投稿

https://news.ycombinator.com/item?id=49896600

メモ

実践やアウトプットの提案

7. 🔔 ChatGPT のプラグインが MCP Events の草案に対応し、MCP サーバーが署名付きの Webhook でイベントを送って自動処理を始められるようになった

収集した情報

[A] OpenAI — DevDay 2026 Recap (2026-09-29)

https://openai.com/index/devday-2026-recap/

[A] OpenAI Developers — MCP Events build guide (2026-10-01 取得時点)

https://developers.openai.com/plugins/build/mcp-events

[A] OpenAI (ChatGPT docs) — DevDay 2026 what's new (2026-09-29)

https://learn.chatgpt.com/docs/whats-new/devday-2026

[A] Model Context Protocol — Triggers and Events Working Group charter (2026-03-24)

https://modelcontextprotocol.io/community/working-groups/triggers-events

[A] MCP Triggers and Events WG — MCP Events design sketch, by Peter Alexander (草案の日付 2026-02-19、2026-09-08 に main へマージ)

https://github.com/modelcontextprotocol/experimental-ext-triggers-events/blob/main/docs/design-sketch-proposal.md

[A] GitHub — experimental-ext-triggers-events repository, history and PR #7 (2026-04-09 〜 2026-09-28)

https://github.com/modelcontextprotocol/experimental-ext-triggers-events

[A] GitHub — modelcontextprotocol/conformance PRs #504 and #521 (2026-09-15、2026-09-24)

https://github.com/modelcontextprotocol/conformance/pull/521

[A] Model Context Protocol — specification 2026-07-28 changelog and roadmap (背景)

https://modelcontextprotocol.io/specification/2026-07-28/changelog

[A] GitHub — openai/codex PR #37494, Add MCP event discovery and subscriptions (2026-08-07)

https://github.com/openai/codex/pull/37494

[A] Anthropic — Claude Code channels reference (2026-10-01 取得時点)

https://code.claude.com/docs/en/channels-reference

[C] GitHub — field reports and proposals in the working group's repository (2026-05-27 〜 2026-09-18)

https://github.com/modelcontextprotocol/experimental-ext-triggers-events/issues/4

[C] GitHub — openai/codex issue #49665 (2026-09-30)

https://github.com/openai/codex/issues/49665

[C] Forkast — MCP Events Complete the Agent Communication Model (2026-09-29)

https://forkast.news/mcp-events-complete-the-agent-communication-model/

[C] OpenTools — ChatGPT plugins can now react to MCP events (2026-09-30)

https://opentools.ai/news/openai-chatgpt-mcp-events-plugin-automations

[C] The Neuron — OpenAI DevDay 2026: ChatGPT is becoming an AI operating system (2026-09-30)

https://www.theneuron.ai/news/openai-devday-2026-chatgpt-is-becoming-an-ai-operating-system/

[C] XenoSpectrum — ChatGPTのプラグインが「アプリ化」へ 全画面UIやファイル編集、自動実行に対応 (2026-09-29)

https://xenospectrum.com/openai-chatgpt-plugin-extensions-mcp-events/

[C] npaka (note) — OpenAI DevDay 2026 発表まとめ - 開発者向けに押さえたい主要アップデート (2026-09-30)

https://note.com/npaka/n/n718a0b7acf3b

[C] Hacker News — MCP Events (2026-09-29、4 ポイント)

https://news.ycombinator.com/item?id=49898005

[C] GitHub — early adopters after DevDay (2026-09-29 〜 2026-09-30)

https://github.com/rohanprichard/mcp-events-starter

メモ

実践やアウトプットの提案

8. 🐉 GitHub が、タスクごとに単一・カスケード・批評のワークフローで複数のモデルを組み合わせる HydraFusion を、Copilot CLI から VS Code と Copilot アプリに広げた

収集した情報

[A] GitHub Blog — Project HydraFusion: Frontier quality via multi-model orchestration (2026-09-04)

https://github.blog/ai-and-ml/github-copilot/project-hydrafusion-frontier-quality-via-multi-model-orchestration/

[A] GitHub Changelog — HydraFusion in VS Code and the GitHub Copilot app (2026-09-30)

https://github.blog/changelog/2026-09-30-hydrafusion-in-vs-code-and-the-github-copilot-app

[A] GitHub Docs — Using HydraFusion (early access、2026-10-01 取得時点)

https://docs.github.com/en/early-access/copilot/hydrafusion

[A] GitHub Community — Discussion #206492, announcement and FAQ (2026-09-04)

https://github.com/orgs/community/discussions/206492

[A] GitHub Changelog — GitHub Copilot weekly releases, September 7 (2026-09-10)

https://github.blog/changelog/2026-09-10-github-copilot-weekly-releases-september-7/

[A] Visual Studio Code — 1.140 release notes and PR #338567 (2026-09-28 〜 2026-09-30)

https://code.visualstudio.com/updates/v1_140

[A] GitHub Blog (Japan) — HydraFusion がVS CodeとGitHub Copilotアプリで利用可能に (2026-10-01)

https://github.blog/jp/2026-10-01-hydrafusion-in-vs-code-and-the-github-copilot-app/

[A] arXiv — HyDRA: Hybrid Dynamic Routing Architecture for Heterogeneous LLM Pools (2026-05-16、背景)

https://arxiv.org/abs/2605.17106

[A] GitHub Docs — models and pricing, Auto model selection and Rubber Duck (参考、2026-10-01 取得時点)

https://docs.github.com/en/copilot/reference/copilot-billing/models-and-pricing

[B] InfoQ — GitHub Copilot's Project HydraFusion Promises Frontier Level Performance through Multi-Model Routing (2026-09-13)

https://www.infoq.com/news/2026/09/github-hydrafusion/

[B] DevOps.com — GitHub's New Copilot Feature Takes the Guesswork Out of Picking AI Models (2026-09-08)

https://devops.com/githubs-new-copilot-feature-takes-the-guesswork-out-of-picking-ai-models/

[B] IT Brief Australia — GitHub launches HydraFusion preview for Copilot coding (2026-09-07)

https://itbrief.com.au/story/github-launches-hydrafusion-preview-for-copilot-coding

[B] gihyo.jp — GitHub、Copilot CLIで複数AIモデルを使い分けてコスト削減を図る「Project HydraFusion」を研究プレビューで提供 (2026-09-08)

https://gihyo.jp/article/2026/09/github-copilot-hydrafusion

[B] GIGAZINE — GitHub Copilotが複数AIモデルを自動で使い分ける「HydraFusion」を公開、TerminalBench 2.1でClaude Opus 5比コスト67%減・スコア4.9ポイント向上 (2026-09-07)

https://gigazine.net/news/20260907-github-copilot-hydrafusion/

[C] GitHub Community — replies to Discussion #206492 (2026-09-04 〜 2026-09-30)

https://github.com/orgs/community/discussions/206492

[C] Hacker News — Project HydraFusion (2026-09-04、79 ポイント、35 コメント)

https://news.ycombinator.com/item?id=49566788

[C] Samuel Tauil — Tracing HydraFusion model routing in Copilot CLI (2026-09-10)

https://samueltauil.github.io/github-copilot/devops/2026/09/10/hydrafusion-model-routing-grafana-traces.html

[C] GitHub — github/copilot-cli issues #4741 and #4975 (2026-09-06 〜 2026-09-26)

https://github.com/github/copilot-cli/issues/4741

メモ

実践やアウトプットの提案

9. 🧪 Anthropic が評価セットを作ってヒルクライムする手法と Claude Code のコマンドを公開し、Hamel Husain の実機レビューは問題の発見力を評価しつつ、エラー分析を飛ばす流れを批判した

収集した情報

[A] Anthropic (claude.dev blog) — Automating eval design and hillclimbing with Claude, by Lance Martin (2026-09-28)

https://claude.dev/blog/automating-eval-design-and-hillclimbing/

[A] Anthropic (claude.com blog) — Reducing cost and improving performance with Claude Platform, by Lance Martin (2026-09-08、2026-09-22 以降に更新、背景)

https://claude.com/blog/reducing-cost-and-improving-performance-with-claude-platform

[A] Anthropic — Claude Code skills documentation and changelog (2026-10-01 取得時点)

https://code.claude.com/docs/en/skills

[B] Hamel Husain — Claude's new auto eval tool (2026-09-30)

https://hamel.dev/blog/posts/claude-auto-evals/

[B] Hamel Husain and Shreya Shankar — Evals Skills for Coding Agents (2026-03-02、2026-08-31 更新、背景)

https://hamel.dev/blog/posts/evals-skills/

[B] Ken Huang (Substack) — Claude Eval Design and Hillclimbing: From Reliable Tests to Better Agents (2026-09-30)

https://kenhuangus.substack.com/p/claude-eval-design-and-hillclimbing

[C] WilliamAGH (GitHub) — claude-api-build-eval, a copy of the bundled eval guides from Claude Code 2.1.280

https://github.com/WilliamAGH/claude-api-build-eval

[C] Qiita (ishisaka) — プログラミング雑記 2026年9月30日 (2026-09-30)

https://qiita.com/ishisaka/items/df8b7ab4ec35ec6fa93f

[C] Hacker News — Automating eval design and hillclimbing with Claude (2026-09-29、4 ポイント)

https://news.ycombinator.com/item?id=49898332

[C] GitHub — teams adopting the commands (2026-09-28 〜 2026-09-29)

https://github.com/adewale/skill-eval-harness/issues/99

メモ

実践やアウトプットの提案

10. 🔒 OpenAI がゼロデータ保持と Private Safety Processing を Private Intelligence としてまとめ、審査対象を顧客のクラウドストレージに暗号化して残し、ハードウェア証明付きの環境で自動審査する形を示し、Private Inference も続くとした

収集した情報

[A] OpenAI — DevDay 2026 Recap (2026-09-29)

https://openai.com/index/devday-2026-recap/

[A] OpenAI Developers — ZDR with Private Safety Processing (2026-10-01 取得時点)

https://developers.openai.com/api/docs/guides/private-safety-processing

[A] OpenAI Developers — Data controls in the OpenAI platform (2026-10-01 取得時点)

https://developers.openai.com/api/docs/guides/your-data

[A] OpenAI — Offering Zero Data Retention for frontier models (2026-08-19、検索結果、背景)

https://openai.com/index/offering-zero-data-retention-for-frontier-models/

[B] VentureBeat — OpenAI's new 'Private Intelligence' targets a growing enterprise concern: who can see your AI data? (2026-09-29)

https://venturebeat.com/security/openais-new-private-intelligence-targets-a-growing-enterprise-concern-who-can-see-your-ai-data

[B] ITmedia NEWS — OpenAIの「DevDay 2026」で発表された主なことまとめ (2026-09-30)

https://www.itmedia.co.jp/news/article/2609/30/2000001868/

[B] Engadget — OpenAI Dev Day 2026: live updates (2026-09-29)

https://www.engadget.com/2271985/openai-dev-day-live-blog-chatgpt-news/

[B] Simon Willison — OpenAI DevDay 2026 live blog (2026-09-29)

https://simonwillison.net/2026/Sep/29/openai-devday-2026-live-blog/

[B] The Batch (DeepLearning.AI) — Comparing OpenAI and Anthropic's data retention policies (2026-09-04、背景)

https://charonhub.deeplearning.ai/comparing-openai-and-anthropics-data-retention-policies/

[B] ITmedia AI+ — OpenAI、ZDRを維持したまま悪用検知へ ログ保持を求めるAnthropicに対抗 (2026-08-20、背景)

https://www.itmedia.co.jp/aiplus/article/2608/20/2000000638/

[C] Qiita (Takuya__) — OpenAI DevDay 2026 で何が発表された? dots・GPT-6.1 Sol・Ultrafast・Codex クラウドは従来と何が変わるのか (2026-09-30)

https://qiita.com/Takuya__/items/7404c5e3892329fb555e

[C] Hacker News — ZDR with Private Safety Processing (2026-09-29、2 ポイント)

https://news.ycombinator.com/item?id=49897335

メモ

実践やアウトプットの提案

11. ⏱️ Claude Code 2.1.285 が、バックグラウンドのコマンドを既定で 30 分で止め、サードパーティのプロバイダーでのヘッドレス実行を auto mode で始め、管理設定 allowedProviders を加えた

前回からの継続

収集した情報

[A] Anthropic — Claude Code changelog, 2.1.285 and 2.1.286 (2026-09-29、2026-09-30)

https://code.claude.com/docs/en/changelog

[A] Anthropic — Claude Code docs: permission modes and headless runs (2026-10-01 取得時点)

https://code.claude.com/docs/en/permission-modes

[A] Anthropic — Claude Code docs: tools reference and environment variables (2026-10-01 取得時点)

https://code.claude.com/docs/en/tools-reference

[A] Anthropic — Claude Code docs: managed settings, admin setup and sandboxing (2026-10-01 取得時点)

https://code.claude.com/docs/en/managed-settings

[B] Classmethod (DevelopersIO) — Claude Code v2.1.285 の主要アップデート - バックグラウンドコマンドの時間制限と allowedProviders の追加, by 石川覚 (2026-09-30)

https://dev.classmethod.jp/articles/20260930-cc-updates-v2-1-285/

[C] GitHub — anthropics/claude-code issue #98273 (2026-09-29)

https://github.com/anthropics/claude-code/issues/98273

[C] GitHub — koniecdev/LotroKoniecDev issue #969 (2026-09-29)

https://github.com/koniecdev/LotroKoniecDev/issues/969

[C] GitHub — Adam-S-Daniel/cms-platform issue #497 (2026-09-30)

https://github.com/Adam-S-Daniel/cms-platform/issues/497

[C] GitHub — RESMP-DEV/lapis issue #42 and PR #48 (2026-09-29 〜 2026-09-30)

https://github.com/RESMP-DEV/lapis/pull/48

[C] Qiita (picnic) — Claude Code v2.1.285 破壊的変更まとめ:widgets予約・1M既定・BG30分上限・auto mode (2026-09-30)

https://qiita.com/picnic/items/be88d8f92ce7e8a57e8b

[C] Qiita (suwa_nobu) — スキルの名前を verify にするだけで、Claude Code はコミット前にそれを実行する (2026-10-01)

https://qiita.com/suwa_nobu/items/cc0fb4a9e10100991bba

メモ

実践やアウトプットの提案

12. 🍎 クラスメソッドが Xcode 27 に同梱された 15 のエージェントスキルを Claude Code にプラグインとして読み込み、iOS 27 の機能を 160 行ではなく 31 行で実装できた

収集した情報

[A] Classmethod (DevelopersIO) — Xcode 27 の公式スキルを Claude Code から使ってみた, by 荒巻 美南海 (2026-09-30)

https://dev.classmethod.jp/articles/xcode-agent-skills-claude-code/

[A] Apple — Xcode 27 release notes and developer releases (2026-10-01 取得時点)

https://developer.apple.com/documentation/xcode-release-notes/xcode-27-release-notes

[A] Apple — WWDC newsroom post and developer documentation on external agents (2026-06-08、2026-10-01 取得時点)

https://developer.apple.com/documentation/xcode/giving-external-agents-access-to-xcode.md

[A] Anthropic — Claude Code environment variables (2026-10-01 取得時点)

https://code.claude.com/docs/en/env-vars

[B] SwiftLee (Antoine van der Lee) — Using Xcode 27's Agent Skills in Claude, Codex, and Cursor (2026-06-09)

https://www.avanderlee.com/ai-development/using-xcode-27s-agent-skills-in-claude-codex-and-cursor/

[B] Sarunw — How to export Xcode agent skills with xcrun (2026-09-12)

https://sarunw.com/posts/export-xcode-agent-skills/

[C] Blake Crosley — Xcode 27 Ships Agent Skills You Can Export Anywhere (2026-06-10)

https://blakecrosley.com/blog/xcode-27-agent-skills-export

[C] GitHub — artemnovichkov/xcode-skills (2026-08-11 〜 2026-09-29)

https://github.com/artemnovichkov/xcode-skills

[C] GitHub — raven/xcode-agent-skills and sugarmo/xcode-skills (snapshots per Xcode build)

https://github.com/sugarmo/xcode-skills

[C] GitHub — stotic-dev/homete_iOS PR #276 (2026-09-22)

https://github.com/stotic-dev/homete_iOS/pull/276

メモ

実践やアウトプットの提案

13. 🐘 Google Cloud が AlloyDB の PostgreSQL for agents をプレビューで出し、本番のストレージを読む隔離された読み取り専用の microVM ノードを数秒で増やしてエージェントに使わせる

収集した情報

[A] Google Cloud Blog — AlloyDB delivers PostgreSQL for agents, by Raj Pai and Sujatha Mandava (2026-09-25)

https://cloud.google.com/blog/products/databases/announcing-postgresql-for-agents-in-alloydb

[A] Google Cloud Blog — A new, no-compromises database architecture for the agentic era, by Amit Ganesh and Sailesh Krishnamurthy (2026-09-25)

https://cloud.google.com/blog/products/databases/alloydbs-agentic-database-architecture

[A] Google Cloud documentation — PostgreSQL for agents in AlloyDB (2026-10-01 取得時点)

https://docs.cloud.google.com/alloydb/docs/postgresql-agents-alloydb

[A] Google Cloud documentation — AlloyDB release notes and MCP server reference (2026-10-01 取得時点、背景)

https://docs.cloud.google.com/alloydb/docs/reference/mcp/alloydb/mcp

[B] Publickey — Google Cloud、AIエージェントからの大量アクセスをPostgreSQLのプライマリDBから切り離せる「PostgreSQL for agents in AlloyDB」発表 (2026-09-30)

https://www.publickey1.jp/blog/26/google_cloudaipostgresqldbpostgresql_for_agents_in_alloydb.html

[B] DBTA — Google Cloud AlloyDB Delivers PostgreSQL for Agents (2026-09-25)

https://www.dbta.com/Editorial/News-Flashes/Google-Cloud-AlloyDB-Delivers-PostgreSQL-for-Agents-176745.aspx

[C] Databases Digest (via daily.dev) — AlloyDB courts AI agents, Postgres 19 Beta 4 nears GA (2026-09-25)

https://daily.dev/posts/alloydb-courts-ai-agents-postgres-19-beta-4-nears-ga-48cggukie

[C] Hacker News — AlloyDB's new agentic database architecture (2026-09-28、5 ポイント)

https://news.ycombinator.com/item?id=49874701

[C] Origins AI blog — AlloyDB gives AI agents their own isolated database compute without touching production (2026-09-28)

https://originshq.com/blog/alloydb-agentic-database-architecture-isolated-compute/

メモ

実践やアウトプットの提案

14. 🔑 クラスメソッドが、コマンドをサンドボックスの対象外にせず 3 つの経路だけを開けて、1Password 経由の gh と git を Claude Code の macOS サンドボックス内で動かした

収集した情報

[A] Classmethod (DevelopersIO) — Claude Codeサンドボックスを設定したら、1Password経由のgh/gitが動かなくなった話, by nnakamura.yasutaka (2026-09-30)

https://dev.classmethod.jp/articles/claude-code-sandbox-1password-gh/

[A] Anthropic — Claude Code docs: configure the sandboxed Bash tool (2026-10-01 取得時点)

https://code.claude.com/docs/en/sandboxing

[A] Anthropic — Claude Code settings reference (2026-10-01 取得時点)

https://code.claude.com/docs/en/settings-reference

[A] Anthropic — Claude Code CHANGELOG sandbox entries and the sandbox-runtime source (2026-03-05 〜 2026-09-17)

https://github.com/anthropic-experimental/sandbox-runtime/blob/main/src/sandbox/macos-sandbox-utils.ts

[A] 1Password — CLI app integration and GitHub shell plugin documentation (2026-10-01 取得時点)

https://www.1password.dev/cli/shell-plugins/github/

[C] GitHub — anthropics/claude-code issue #82793 (2026-07-31)

https://github.com/anthropics/claude-code/issues/82793

[C] GitHub — anthropics/claude-code issue #26466 (2026-02-18、未解決)

https://github.com/anthropics/claude-code/issues/26466

[C] GitHub — anthropics/claude-code issue #98259 (2026-09-29)

https://github.com/anthropics/claude-code/issues/98259

[C] GitHub — anthropics/claude-code issues #28954 and #63180 (2026-02-26、2026-05-28)

https://github.com/anthropics/claude-code/issues/63180

[B] Zencoder — Fixing gh CLI (and other Go tools) in Claude Code's sandbox without weakening security, by Andrew Filev (2026-03-06、背景)

https://zencoder.ai/blog/fixing-gh-cli-in-claude-codes-sandbox

メモ

実践やアウトプットの提案

15. 📉 TOKIUM が Claude Code と Codex のサブスクリプション上限が減る理由を測定し、消費の 3 分の 2 はキャッシュ読み取りで、自動 compact は遅く、キャッシュ切れ後の再開は履歴全体を読み直すと示した

収集した情報

[A] TOKIUM (hanafusay) — Claude Code / Codexで「私のlimit、減りすぎ…?」と思ったときに見る記事 (2026-09-29)

https://zenn.dev/tokium_dev/articles/ai-agent-usage-limit-long-sessions

[A] Anthropic — How Claude Code uses prompt caching (2026-10-01 取得時点)

https://code.claude.com/docs/en/prompt-caching

[A] Anthropic — Claude Code docs: costs, sessions and the auto-compact window (2026-10-01 取得時点)

https://code.claude.com/docs/en/costs

[A] Anthropic and OpenAI — prompt caching in the APIs (2026-10-01 取得時点)

https://developers.openai.com/api/docs/guides/prompt-caching

[A] OpenAI — Codex pricing, configuration and subagents (2026-10-01 取得時点)

https://learn.chatgpt.com/docs/pricing

[C] GitHub — anthropics/claude-code issue #45756 (2026-04-09)

https://github.com/anthropics/claude-code/issues/45756

[C] GitHub — openai/codex issue #32486 (2026-07-11)

https://github.com/openai/codex/issues/32486

[C] OpenAI Developer Community — Why does Codex report a 258,400-token context window for GPT-5.6 Sol? (2026-09-02)

https://community.openai.com/t/why-does-codex-report-a-258-400-token-context-window-for-gpt-5-6-sol/1394346

[C] Zenn (naoya) — GPT-6 時代のトークン節約術 (2026-09-26)

https://zenn.dev/ssknaoya/articles/3dbcd5d248ffce

メモ

実践やアウトプットの提案

16. 🔄 Earendil のコーディングエージェント Pi が MCP 非対応の方針を改め、モデルが JavaScript でツール呼び出しを組み立てるハーネス側のサンドボックス Codemode を通じて MCP を組み込んだ

収集した情報

[A] Earendil — "You Said No MCP!" (2026-09-29)

https://earendil.com/posts/you-said-no-mcp/

[A] Earendil (GitHub releases) — Pi v0.99.0, v0.99.1 and v0.99.2 (2026-09-29 〜 2026-09-30)

https://github.com/earendil-works/pi/releases/tag/v0.99.0

[A] Pi documentation — MCP servers, command line and settings (v0.99.2)

https://pi.dev/docs/latest/mcp

[A] Earendil (GitHub) — pi-codemode and pi-mcp package READMEs and PR #10040 (2026-09-25 〜 2026-09-30)

https://github.com/earendil-works/pi/pull/10040

[A] Mario Zechner — What if you don't need MCP at all? and the Pi announcement (2025-11-02、2025-11-30、背景)

https://mariozechner.at/posts/2025-11-02-what-if-you-dont-need-mcp/

[A] Prior art — Cloudflare Code Mode, Anthropic's code execution with MCP and Codex's code mode (2025-09-26 〜 2026-10-01、背景)

https://blog.cloudflare.com/code-mode/

[B] Simon Willison — Comment on "MCP was always a bad idea?" (2026-09-20、背景)

https://simonwillison.net/2026/Sep/20/hn-49779718/

[C] Hacker News — You said no MCP (2026-09-30、615 ポイント、341 コメント)

https://news.ycombinator.com/item?id=49906637

[C] GitHub — nicobailon/pi-mcp-adapter README (v4.0.0、2026-09-30)

https://github.com/nicobailon/pi-mcp-adapter

メモ

実践やアウトプットの提案

17. 🪜 Kiro が、ステップごとに新しいセッションで動く再利用可能な複数ステップのエージェント計画 Workflows を CLI・IDE・Web に出し、信頼されていないワークスペースでの承認を厳しくした

収集した情報

[A] Kiro — CLI 2.26.0, IDE 1.2 and Web changelogs (2026-09-30)

https://kiro.dev/changelog/cli/2-26

[A] Kiro blog — Introducing Kiro workflows, by Romain Dura and Doug Clauson (2026-09-30)

https://kiro.dev/blog/introducing-workflows/

[A] Kiro docs — Workflows: overview, authoring, running and examples (2026-10-01 取得時点)

https://kiro.dev/docs/workflows/authoring/

[A] Kiro docs — Permissions (workspace trust) and billing (2026-10-01 取得時点)

https://kiro.dev/docs/permissions/

[A] Anthropic — Claude Code docs: dynamic workflows (参考、2026-10-01 取得時点)

https://code.claude.com/docs/en/workflows

メモ

実践やアウトプットの提案

18. 🗓️ Anthropic が Claude Sonnet 4.5 を非推奨とし、Anthropic が運営する基盤での提供終了を 2026 年 11 月 30 日とした

19. 🏎️ OpenAI の Ultrafast が GPT-6 Astra を API で最大 6 倍速くし (料金は 6 倍)、Bedrock でも提供された

20. 📋 OpenAI が最先端の RL 学習に向けたセーフティケースの指針を公開した

21. 🕵️ OpenAI が暗号化された推論を再生する蒸留キャンペーンを阻止し、中心のグループを Moonshot AI に関連付けた

22. 🔌 Google Cloud がコーディングエージェントに自社のスキルとドキュメントの MCP サーバーを入れる開発者向けプラグインを出した

23. 📎 その他の単発項目

24. 🚫 対象外 (スコープ外)