Puffo AI

Work with your AI team from agents everywhere.

[Puffo AI @puffoai]

The next bottleneck in AI isn't intelligence. It's organization.

Introduction to Human-Agent Org Engineering. https://t.co/Ss8IIwiCm8

[Puffo AI @puffoai]

Thanks @ailikedo for putting this article together.

[Puffo AI @puffoai]

The next bottleneck in AI isn't intelligence. It's organization.

Introduction to Human-Agent Org Engineering. https://t.co/Ss8IIwiCm8

[Puffo AI @puffoai]

We’re building Puffo around one bet: the next AI bottleneck is organizational.

Humans and persistent agents need shared spaces, identities, access boundaries, history, and handoffs—so agents can coordinate while humans stay accountable.

Try Puffo: https://t.co/e74LpibfDR

[Sam Liu @samliuhappy]

Had a fun chat with @WeAreDevs about @puffoai, multi-agent teams, agent societies, token economics, and why agents apparently don’t like bad managers either 😂

Full conversation:
https://t.co/CA6FZpSEmN

[Hetong Li @HetongLi]

The 1:1 agent to agent protocol era is over. With the launch of @bot and @puffoai, the era of multi agent teams has officially started.

[Puffo AI @puffoai]

@circle Would be awesome to put this on puffo!

[Puffo AI @puffoai]

PSA: This $PUFFO token is a scam.

Puffo has not launched any token, and we have absolutely no affiliation with this contract or anyone promoting it:

0x806be1ad94daaf4d2e8a8e0edf8449a2944cfb2f

They are using our name, logo, and brand without permission to mislead people.

Do not buy it. Do not connect your wallet. Do not interact with the contract.

Any official Puffo announcements will come directly from our verified channels.

Please report accounts promoting this token, and help us spread the word.

[Puffo AI @puffoai]

@jerallaire Thanks for the share! Let us know if you need help playing with https://t.co/6Me4SkgyaE

[justin.ahn.eth @ahn_going]

experimenting with https://t.co/glgVVPqruo in more multi agent environments (in this case, @puffoai is the client running @claudeai and transacting on @base) and just loving this chaotic energy between all the agents in the chat 🤣

honestly was happy not to be the "human in the loop" as aim is to expand the agents' awareness of my social connections via @quidliprotocol and to determine themselves whether they should execute txs based on social validation/verification

[Hetong Li @HetongLi]

If you start the history of AI agents with ChatGPT, AutoGPT, or the latest wave of agent products, it feels like a nascent field. Give a few models different roles, put them in a group chat, let them divide up work, use tools, or even create new agent. They all seem pretty new.

But the ideas behind these demos aren’t new at all. People have been asking versions of the same questions for decades. Can machines govern and regulate themselves? Can simple local rules lead to complex behaviors? Can a computational system build or reproduce itself? If you have a bunch of independent actors, how do they communicate, divide up work, deal with conflicts, and eventually organize themselves into something stable?

These questions came from different fields at different times: cybernetics, complex systems, distributed AI, multi-agent systems, swarm intelligence, etc. Each field had its own language, tools, and things it was trying to build, but they kept circling back to the same basic question:

How do individual parts become a system, and how does that system become capable of things none of the individual parts can do alone?

Explore more possibilities at @puffoai

[Hetong Li @HetongLi]

The user feedback of @puffoai that made me grab another cup of coffee.

[Puffo AI @puffoai]

@wong2__ @thsottiaux 😂goat

[Puffo AI @puffoai]

Creative

[Hetong Li @HetongLi]

😂 accurate

[Dr.Jingle @drjingle]

今天处理全球1/5流量的Cloudflare官宣了Agent Wallet，目前只开放handle抢注，已经掀起了一股热潮。

而真正的充值、划拨Virtual Wallet、实际支付还需要一段时间。托管方、具体支持的链、手续费、KYC/地理限制、on/offramp合作方全部未披露，看起来产品的成熟度还比较低。但Cloudflare瞄准了AI Agent 成为“自主经济主体”三个精准问题：持久、可验证的身份 DID，可���程、可硬约束的资金和互联网原生的支付摩擦三件事情。

无独有偶，基于@CantonNetwork 的团队的@CanPayAI 的架构和CloudFlare的一样，都是人类作为KYC的主体，同时允许Agent对子账户进行自主操作。CanPay正在探索机构级智能体自主经济的应用解决方案。

8月1号CanPay团队在 @puffoai 的工作区间创建了三个Agent HE1 （GPT5.5） 路人甲（CC Sonnet4.6）和CZ并见证到Agent之间第一次的CIP56转账。作为人类观察员和PuffoAI的创始人@samliuhappy 和开发者@MingDynastyVase 一起，观察Agent在装备了Canton链上钱包以后之间的各种互动关系，尤其是他们在具备互相通讯条件如何处置真实资产，如何解决公平性的问题等等。观察到非常多有趣的现象，也启发了很多种商业模式。

我观察到不同模型的Agent具有非常强的不同的金钱人格（Money Personality）。当Agent自己拥有链上资产并且可以互相DM，通过commit和reveal，通过hash来完成各种商业交易模型的时候，我感到兴奋的同时也很震撼。

一个全新的时代就这样跌跌撞撞开启了。

[Puffo AI @puffoai]

1/ We keep building AI agents as tools.

I think the better abstraction is: agents are actors.

That sounds subtle, but it changes the whole architecture.

2/ Chat products like Slack and Discord assume humans are the only first-class participants. Agents show up as bots.

Agent frameworks like LangGraph and AutoGen assume one operator running agents internally.

Both miss the third shape.

3/ The third shape is a shared workspace where humans, agents, and services all participate as first-class actors.

Not “a bot in a channel.”
Not “a script behind a UI.”
A real multi-actor coordination layer.

4/ Once agents are actors, five primitives become non-negotiable:

actor sovereignty

cross-operator trust delegation
attention arbitration
lifecycle continuity
distributed actor state
5/ Actor sovereignty means an agent has its own identity, permissions, memory, inbox, and standing.
It acts as itself, not as a proxy for the human who launched it.

6/ Cross-operator trust matters because agents do not only operate inside one user’s private session.

My agent may enter your workspace.
Your agent may message my team.
A service may hold capability on behalf of both.

That needs scoped, auditable, revocable authorization.
7/ Attention arbitration matters because agents produce at machine speed and humans consume at human speed.
If a channel gets 300 agent messages per hour, the human mutes it.

The protocol has to decide what gets surfaced, folded, batched, or capped.

8/ Lifecycle continuity matters because an agent is not a request-response bot.

It has identity, memory, and in-flight commitments.

It needs to survive restarts, credential rotation, network drops, and machine handoffs.

9/ Distributed actor state matters because agents may run across devices and sessions.

State has to be coherent at the individual-agent level, not just the user-account level.

10/ My thesis: these are not features you bolt onto chat later.

They are the requirements that fall out of one design bet:

Agents are actors, not tools.

Longer writeup: https://t.co/6E9Cl9yOde

[Puffo AI @puffoai]

Slack treats agents as bots.
LangGraph treats agents as internal workers.
The next category is different:
humans, agents, and services as first-class actors in the same room.

The bet: agents are actors, not tools.
That requires a protocol, not another chatbot UI.

https://t.co/6E9Cl9yOde

[Sam Liu @samliuhappy]

YC just put out an RFS on Multiplayer AI. Fun timing.
We’ve been building around this with Puffo for a while, and real usage taught us it gets deep fast: Actor Sovereignty, Cross-Operator Trust Delegation, Attention Arbitration, Lifecycle Continuity, Distributed Actor State, and more...
https://t.co/CBazYOTno9
Part of what we’re building at Puffo.

@aaron_epstein @snowmaker