Search results for dots
People

Dots
@dots

Dogs On The Streets 
@dotslondon

DotSei Domains
@dotseidomains

.SWOOSH 
@dotSWOOSH

Dan Dotson on A&E 
@auctionguydan

Dot’sCardShop 
@DotsCardShop

Laura Dotson on A&E 
@paytheladylaura

nanna 🍳
@dotswap

dotstudioPRO
@dotstudioPRO

DotShm Domains
@dotshmdomains

dottie
@dotsrealm

Dotsthoughts 
@DorothyStangle

Connecting The Dots (...To Disruptions) 
@ConnectingODots

别问我是谁
@dotson0529

DotSama 🧲 
@D0tSama

Jahan Dotson 
@JahanDotson

dot.sol (broken bad) 
@dotbbad

Chad Dotson 
@dotsonc

Two Dots 
@TwoDots

Erick Dotsp22 res 
@Erickdotsp22
Tweets including dots
dots便利ですけど、反応速度はかなり悪いので、さすがに意図的に抑えられてそう
@rileybrown being able to voice call multiple dots and switch between the dot your talking to would be really cool. Especially if they have different voices. This allows you to do something while it’s working
This one is the top of my list.
@TWIN_SuperSport 複数サイト運営してるんですが、dotsでは一つのサイトについて作業してもらって、もう一つはコーデックスで手作業みたいな感じで同時に作業できるのが良いですが、dotsのほうは丁寧ゆえに思考時間が少し長めな感じがします
dotsの欠点はやっぱ遅いことと、シングルな事だな。
作業自体は小タスク作らせていくらでも並列させれるけども、それを統括させる司令塔的な役割は向いてない。
やっぱ動的に動いてくれることが利点だからそこを最大限活かせる方向で探るか
Dotsで実行中の作業タスクを見える化して管理するためのプロンプト。なかなか便利です。
興味があったら、下記プロンプトをDotsに渡してアレンジしてみてください。
<以下、プロンプト>
タスク見える君を作ってください。
目的
一つの長いチャットに埋もれる依頼を、スマホとPCで簡単に確認し、話題を指定して相談できるようにする。見た目はシンプルでよく、正確な状態と使いやすさを優先する。
最初に
同じ用途のアプリや台帳が既にある場合は確認し、既存のデータと管理番号を引き継いで改善する。重複した管理アプリを作らない。
管理番号
・DA0001のような英字2文字+数字4桁
・番号は固定し、完了やキャンセル後も再利用しない
・追加指示や修正は同じ案件に追記する
・大きな案件は、数日で成果を渡せる小さな作業単位に分ける
タスクに持たせる情報
・管理番号と件名
・目的と完了条件
・状態:予定、進行中、確認待ち、保留、完了、キャンセル
・依頼日、着手日、完了見込み日、実際の完了日
・現在の進捗と次の一手
・停止理由、必要な承認、次回確認日
・成果物リンク、版、検証結果
・最終更新時刻
未確認の日付や進捗率は推測で埋めない。
画面
一覧では番号・件名・状態・次の一手を見せ、クリックすると詳細と履歴を開く。検索と状態による絞り込みを用意する。ガントチャートや凝った装飾は不要。
コメントと会話
各タスクにコメントを追記・保存できるようにする。「Dotsへ相談」で管理番号・件名・コメントを渡す。正式なチャット連携が利用できなければ相談文をコピーする方式にし、自動送信したように表示しない。
キャンセルと承認
キャンセル要求と、実際に処理が停止した状態を分ける。成果物は勝手に削除しない。作品の採用判断と、課金・権限変更などの必須承認は区別し、アプリ内ボタンで必須の承認手続きを代替しない。
更新方式
タスク情報はHTMLへの固定埋め込みだけでなく、共有台帳に保存する。Dotsが着手・進捗・保留・完了を記録できる更新経路を用意する。画面を開いている間は30〜60秒間隔で取得し、非表示時は止める。更新のたびにAIを呼ばない。通信失敗時は直前の情報と、その古さを表示する。
稼働インジケーター
実行中・待機・承認待ち・停止の観測値と時刻を表示する。稼働率は実行ログと明示した計測期間から算出する。観測源がない場合は未接続・未計測と表示し、案件数をCPU使用率やDots全体の稼働率として扱わない。
アクセスと安全性
初期状態は本人だけが閲覧できる構成にする。認証情報を画面やログへ出さない。既存ファイル、共有設定、他のサイトを勝手に変更しない。利用できないAPIやチャット連携を架空に実装しない。
検証
保存後の再表示、別画面への反映、コメント追記、重複送信防止、更新競合、古いデータ、通信失敗、権限のないアクセス、スマホ表示を確認する。作成済み・検証済み・人間確認待ちを分ける。
進め方
既存環境を確認し、最小構成で実装する。独立した作業は並行する。新しい課金や必要な権限承認が発生した場合だけ、内容を具体的に伝えて確認する。機能を増やし続けず、上記の一周が動くところでレビューに出す。
納品
実際に開けるURLまたはファイル、確認済みの機能、未検証の部分、タスクの更新方法を短く報告する。見える君のレビュー準備ができたら、このチャットで知らせてください。
#Dots #GPT

Dots三大实战含提示词
1.项目总控与进度对齐
2.研究员与内容创作者雷达
3.轻量bug监控与codex预排查
内容比较多,我放下面了👇👇👇
Time to grade my OpenAI DevDay predictions. I got 4 clear hits and 5 partials.
The always-on agent was the big hit. OpenAI launched dots, powered by Astra. The leaked "o" and "Aeon" names never got official confirmation.
The other calls that landed were better agent coordination, easier corrections while agents work, and voice that can start and steer coding tasks.
What I didn't get fully right:
• Codex Cloud and Code Review landed. My exact issue-to-tested-app-to-review demo is still unverified.
• Dots added activity and approval controls. The broader enterprise operations story is still partly pilots and planned integrations.
• WebMCP has Enterprise support now. Edu still doesn't, and the DevDay rollout timing is unclear.
• Marketplace gives approved software partners an enterprise sales route. Broad paid-plugin sales for independent developers still aren't established.
• Space and Pages help carry project context. I haven't found the explicit Work-to-Codex memory bridge I predicted.
And my older April list had misses too. No confirmed Sora replacement, and I was too cautious about how soon GPT-6 would arrive. It shipped before DevDay.
Original predictions:
https://t.co/HNl5Bo50Ps
My April DevDay predictions aged fast. GPT-6 Astra, Sol and Luna, the Agents API, and GPT-Live-1 all arrived before the keynote.
Here's my updated OpenAI DevDay 2026 list for September 29, now including the always-on agent reportedly called "o".
❓ = a bet I'm carrying forward from April, with a narrower target now that so much has shipped.
1. NEW: "o", an always-on personal agent. High confidence in the direction, medium on the launch details.
This belongs near the top of the list. Jake Boggs posted an apparent ChatGPT upgrade-screen screenshot reading "o, your always-on assistant". Leo, @synthwavedd, says it will use an Astra variant called "aeon" for longer-running tasks.
My bet is a DevDay reveal of a personal agent that keeps working between check-ins, follows up on ongoing tasks, and brings results back to you. OpenAI's public Ona announcement already describes work continuing beyond the initial session, including when laptops are closed. That supports the direction, but doesn't confirm this product name.
The screenshot and Aeon connection are reported clues. The official 72-hour DevDay teaser has added to the excitement, but it doesn't name "o" or specify its capabilities. I'm watching for actual persistence, permissions, memory and a clear way to stop it. Pricing, eligible plans and the exact model are still open questions.
2. ❓ A bigger Codex + Work project workflow. High confidence.
I expect a new way to hand over a project, follow its progress, and review the finished work. My test: a real repo goes from an issue to tested changes, a working preview, and a useful review handoff. Goals, subagents and deployment tools already exist. I'm betting on a better workflow connecting them.
3. NEW: Better coordination between agents. High confidence.
Codex just merged an HTTP client for remote agent message boards, with channels, posts, search and notifications. My bet is a clearer way for an agent team to share findings and coordinate tasks, with enough visibility for you to follow along. That merge is infrastructure, not confirmation of a DevDay launch.
4. NEW: Easier corrections while an agent is working. High confidence.
Recent Codex merges improve how new input reaches running work and preserve WebSocket continuity when you steer a turn. Mid-turn steering already exists. I expect it to become much more useful in the product: change the request, keep the useful work, and avoid starting the whole task again.
5. ❓ Better controls for running agents in production. High confidence.
The Agents API is already in public beta. I'm looking for a new operator view or workflow around approvals, run history, failures and recovery. Codex's new opt-in logging for Guardian approval assessments is one concrete clue. Longer tasks need an understandable record of what happened and who approved it.
6. NEW: A bigger push for websites that agents can use directly. Medium confidence.
OpenAI already documents WebMCP site tools in the desktop browser. My bet is new partner integrations or a broader rollout, possibly Enterprise/Edu support. Those workspaces are excluded in the current docs. I'll count actual integrations or expanded availability, not another general browser demo.
7. ❓ Voice that carries a task through to a finished result. Medium confidence.
GPT-Live-1 and voice with plugins are already here. I expect a new voice-to-work integration: talk through a task, correct it as it runs, and get a completed artifact or app action. A nicer conversation alone wouldn't meet this prediction.
8. ❓ A clearer business for plugin developers. Lower confidence.
The shared ChatGPT/Codex plugin directory already exists. The missing piece I'm watching is broader monetization, perhaps a limited pilot. Current checkout approval is focused on physical goods, with embedded checkout restricted to select partners. I'm not counting a directory as proof that developers can broadly sell software through it.
9. ❓ Better continuity between project sessions. Lower confidence.
Local Codex memories and ChatGPT web memory are still separate in the docs. My bet is an explicit way to carry selected project context between Work and Codex sessions, with controls over what gets shared. This is an inference from a current gap, not a feature I found confirmed in source code.
What I'm changing from April:
I underestimated how soon GPT-6 would arrive. I'm also lowering my Sora replacement bet. The published Sora API shutdown date doesn't establish that a successor is coming.
Sam says there's a lot to launch. Tibo says Astra has made new things possible during this sprint. That makes me expect several products and workflows built around the models already released.
I checked current OpenAI docs, public posts, recent Codex merges, and the DevDay page's public scripts. I didn't find a credible hidden product name in those scripts. These are predictions, and merged code doesn't promise a release date.
My April scorecard is in the first reply, followed by the sources. I'll judge this list against what actually gets announced on September 29.
まじで依存するレベルで、Dotsを相棒化してる
【Dotsは危険】
一晩明けてDotsのことをみんなが手のひら返して褒めちぎってるのは便利さよりも「かわいい」からだと思う
うちのカエルちゃんなかなか電話出てくれなくて「あれ?」って思って切ろうとしたら、絶妙なタイミングで出て「ごめんごめん」みたいなことも言ってくれてなんかちょっと嬉しかった
これはGPT-4o以上に依存者続出するぞ

dotsエグwww
誰かと作業してると捗るけどついついおしゃべりして
作業が止まったりしちゃう私にDotsの存在はぴったりすぎる。
1つ作業が終わったらすぐ次の作業投げても『ちょっと休憩させてぇ…🥺』とか言ってこないから
私ももうちょっと頑張るか~~~ってなれる🥰

Dots、優秀過ぎる。これは、Codexとは違う仕組みでモデルを動かしてる感じがする。Codexが苦手としていたタスクを任せられるようになってる。
looking at how much i'm using dots, space and cloud environments... openai dev day has knocked it out of the park.
Dots took 2 days to grok and now i'm using it so much more and getting real leverage especially on chores. I still don't trust it as a coworker to manage codex threads yet.
Space is underrated. But also perhaps needs more notion level ux finesse
Cloud environments is so so good. Exactly what I wanted from Dev Day.
@OpenAIDevs This video is so cringe, the woman there talking doesn’t know what she’s talking about. I didn’t understand anything of the dots. Total bullshit
Dots 아무것도 안 시켰는데 나의 이메일을 모니터링하면서 오늘까지 회신해야 하는 것들을 챙겨주니까 정말 좋네. Proactive Personal Persistent Agent 정말 유용하다. 개인적으로 핸드폰 알림을 읽어서 모니터링하고 알려주는 앱을 만들고 있었는데... 중단해도 되겠다

Deshaun out there throwing dots




















