tokensflow

[tokensflow @tokensflow]

@Xudong07452910 这个也很像opus的风格

[tokensflow @tokensflow]

@YuLin807 整个宇宙在为你闪烁。

[tokensflow @tokensflow]

@0TwT012 那是他杀你子孙了属于是

[tokensflow @tokensflow]

@NotionHelp Hey I wanna use fable but I could not find the model in page and I am business plan,I also not find the model in the workspace settings thx

[tokensflow @tokensflow]

@yusuke_tony 真奇妙，你引用了引用了你的引用

[tokensflow @tokensflow]

@NotionHQ hey where is my fable，is it not for all business user?I am in business plan。

[tokensflow @tokensflow]

@oasisfeng https://t.co/ykT3X5f1Kp试试这种呢

[Diego | AI 🚀 - e/acc @diegocabezas01]

Use Fable 5 as orchestrator and Opus + Codex to execute (to save fable usage):

Fable 5 (max reasoning) = orchestrator
Opus = deep reasoning subagent
Sonnet = mechanical work subagent
Codex = peer Sr. engineer, different perspective

Setup:
1. Set Fable 5 as your main model In Claude Code: /model → Fable 5 → reasoning /effort to max

2. Create 2 subagents with /agents In Claude Code:
deep-reasoner → pinned to opus "Use for reasoning-heavy phases, architecture, debugging complex issues, algorithm design. Think thoroughly, return a concise conclusion the orchestrator can act on."

fast-worker → pinned to sonnet "Use for mechanical tasks, boilerplate, tests, formatting, simple edits. Execute efficiently."

3. Add OpenAI's official Codex plugin (install codex cli in your computer first), In Claude Code type:
/plugin marketplace add openai/codex-plugin-cc
/plugin install codex@openai-codex
/codex:setup

4. Drop this in your CLAUDE.md in your folder:

## Orchestration workflow
You (Fable) are the orchestrator. Plan, decompose, synthesize.
Reasoning-heavy phases → deep-reasoner
Mechanical work → fast-worker
Codex (/codex:rescue --background) is a cracked engineer on par with deep-reasoner, from a different perspective. Treat as a peer, not a reviewer.
High-stakes decisions: task Opus + Codex on the same problem in parallel, synthesize the best of both, without showing either the other's answer. Keep your own context lean.

5. Then prompt Fable like a tech lead: "Goal: [what you want] Context: [files, constraints] You're the lead. Delegate reasoning to deep-reasoner, grunt work to fast-worker, fresh-perspective problems to Codex. Show me your plan first, then execute."

That's it.

[tokensflow @tokensflow]

@SonicAIWizard 负负得���

[tokensflow @tokensflow]

@scaling01 how much the human model per M token

[tokensflow @tokensflow]

@cnzoecomeback

[tokensflow @tokensflow]

气笑了

[Thariq @trq212]

Have seen some questions about the updated classifiers and wanted to clarify.

As with the original classifiers, a small fraction of routine coding and debugging tasks will be flagged and fall back to Opus.

We're excited for guys to get access back tomorrow.

[Mr PitBull Stories @MrPitbull07]

In 1872, an eighteen-year-old girl arrived in a remote gold mining camp in Idaho. Legally, she was not even considered a human being. She did not have money, she did not speak a word of English.

She had been sold by her starving family in China during a famine, smuggled across the ocean, and bought by a wealthy Chinese saloon owner to work as a slave.

Her birth name was Lalu Nathoy, but the miners quickly gave her a simpler name: Polly.

The odds against Polly were overwhelming. Under American law at the time, she was invisible. The local government considered her presence illegal, and the men in the camp viewed her as property.

But Polly possessed a quiet determination that no one saw coming.

While she spent long, exhausting hours scrubbing heavy canvas pants on a washboard, she listened. She memorized every word spoken around her.

She learned English in complete silence, without anyone realizing it until it was too late to stop her.

Polly looked at the rugged miners around her and noticed something crucial.

They had gold in their pockets, but absolutely nothing else.

They had no one to feed them properly, no one to nurse them when they fell ill, and no one to make the brutal survival in the canyon bearable.

She saw a massive void in the market and decided to fill it.

She started cooking, sewing, and providing basic medical care. Every single coin she earned from these side jobs went straight into the dirt floor underneath her bed.

While the men squandered their fortunes on gambling and alcohol, Polly was buying something much more permanent: her independence.

Her life took a dramatic turn when a local saloon keeper named Charlie Bemis was shot in the face during a gambling dispute.

The camp doctor took one look at the horrific wound and declared him a dead man. Polly refused to accept that. She boiled water, sterilized a common crochet hook, and spent hours carefully extracting the bullet from Charlie’s skull.

Against all medical logic, Charlie lived.

Eventually, Polly and Charlie left the mining camp together and moved to Hells Canyon, the deepest river gorge in North America.

The Snake River cut through granite walls so steep that sunlight only hit the valley floor for a few hours a day. On a piece of land that seemed impossible to cultivate, Polly planted a fruit orchard.

She grew cherry and apple trees against the harsh rock cliffs.

When miners down the river got sick with fever or suffered terrible injuries, Polly took them in, becoming the ultimate healer of the canyon.

But her greatest battle was yet to come. In 1892, the U.S. government passed the Geary Act, a harsh law requiring all Chinese residents to carry certificates of residence or face immediate deportation. Polly had no papers.

A federal official traveled down into the canyon specifically to deport her. But when he arrived, he saw the thriving orchard, the vegetable gardens, and the sick men Polly was actively nursing back to health.

Realizing she was the backbone of the entire canyon community, the officer sat at her table, filled out the residence paperwork, and signed it as her witness instead of arresting her.

Polly Bemis lived in her canyon until her death in 1933 at the age of eighty. Today, her cabin is protected as a National Historic Site, and the cherry trees she planted still bear fruit.

Polly Bemis proved that when your spirit is strong enough, human law becomes nothing more than a suggestion.

She began her life in America with absolutely nothing, yet she chose to fill the harsh canyon with sweet fruit, warm meals, and a safe place for people who had no one else to care for them.

Enslaved, isolated, and stripped of every legal right, Polly faced a harsh wilderness and an even harsher society with absolutely no fear. She chose to fight back not with malice, but by building a life of profound purpose and protecting those around her.

[tokensflow @tokensflow]

@AdrianPunk115 我有个长聊天，都是 继续，继续，继续😂

[tokensflow @tokensflow]

@BruceGuai 已转发，求测试

[Bruce Guai @BruceGuai]

这绝对是你当下最值得尝试的 Agent 产品，比肩甚至超越当前 cc 和 codex，在公开评测集中占据 SOTA。其领先的 harness 框架和长任务及loop的表现，和我们独创的 agentic department and okr 系统，能够满足用户的超长期复杂任务，并自主迭代。

经过一周的公测，Matrix 正式和大家见面了。在一周的公测期间，有上万个 0 人 Agent 公司成功创建，当中前列公司已经可以真实的获得收益，并且持续增长。

今天每个人都可以在 MacOS上下载并且构建你的第一个 Agent 公司。支持任意模型 API 驱动，你也可以使用自己的 Codex 或者 Claudecode 订阅套餐来驱动 matrix。使用matrix的订阅会获得更多功能���用权力，以及享受模型折扣。

关注➕转发我们的帖子，为前100位用户送出一笔Matrix的10个会员，当做你的第一个agent公司的初始基金。

[tokensflow @tokensflow]

@Khazix0918 4.8会持续调用工具失败，就很难绷

[tokensflow @tokensflow]

Claude Code weekly limits up 50%
through 7/13, 6pm

[tokensflow @tokensflow]

@Alex_perception 还有一部分原因是电力供应不足

[tokensflow @tokensflow]

https://t.co/NPOeKONbgo

[tokensflow @tokensflow]

开源了一个带侧栏状态的终端。
最大的原则是是克制，好看，轻量。

大小只有30m。
不做大而全，不给认知增负。

最大的特色是侧栏，显示各个目录下的当前agent 状态。
右侧是diff，你可以看到文件改动。

界面很简约好看，动效也很棒，希望喜欢。

不做的清单：
不做��置 AI 聊天、模型接入、MCP 编排。
不做 agent 启动器。
不做 commit / push 写操作。
不做插件、自研渲染、递归分栏。
不做遥测。

标准很简单:让终端继续是终端,而不是变成下一个 IDE,或者下一个 agent 控制台。