
community @GoKiteAI | I am my own Alpha | Kite passport: https://nitter.gq/t.co/lPxgVrWHl8
Joined March 2025
- Tweets10.6K
- Following608
- Followers924
- Likes26.5K
Soo… good afternoon guys 👋
Over the past 3 days, I’ve been working on setting up the @GoKiteAI Agent Passport.
And I can finally say… I’m done 😭
It honestly wasn’t easy fr.
What I thought would be a quick setup turned into a full learning experience. From getting everything running properly in Termux, configuring the agent environment, setting up the passport, understanding permissions, execution flow, and fixing random issues along the way… it took way more patience than I expected.
There were moments I got stuck badly 😭
Moments I had to restart processes.
Moments I kept rechecking the docs just to make sure I didn’t miss something small.
But eventually everything started clicking together.
And once the setup was finally complete, I decided not to stop at “setup successful.”
I wanted to test something real.
So I used the agent to order a book from @amazon using the @crossmint payment system.
And honestly… this was the moment the whole idea of AI agents started feeling real to me.
The process was actually crazy to watch.
First, I defined the task clearly for the agent and configured the permissions inside the Agent Passport so it had controlled access to execute actions safely.
Then I launched the workflow through Termux and let the agent handle the execution flow.
From there, the agent:
•searched for the book using the query system inside the agent workspace
•identified a valid Amazon listing
•prepared the checkout flow
•and completed payment through Crossmint
What stood out to me most was that I wasn’t manually controlling every step.
Once the boundaries and permissions were defined, the agent simply operated within those limits and completed the task.
That balance between:
•autonomy
•permissions
•and real-world execution
is what makes this whole system interesting.
Crossmint acted as the payment bridge connecting the agent’s workflow to an actual real-world purchase, and seeing the order go through successfully honestly changed how I look at AI agents.
It stopped feeling like “AI tools.”
And started feeling more like delegating tasks to a digital operator that can actually execute actions on your behalf.
This is where things start getting really interesting for me 🪁⚡
When running yours, if you run into issues or have questions, feel free to ask me or @0xLaughing.
Shout out to people who helped me
@KingTowdah
@danielnwabuisi0
@marurlemo
@david_grammy
@Kite_Frens_Eco
The Kite × @bitrefill joint campaign is live.
From now until Nov 24, let your agent shop on Bitrefill with Kite Agent Passport and get paid back in $KITE, from a $1,000 reward pool funded by Kite:
▷ 3% back in $KITE on eligible orders, up to $5 per user.
▷ Up to 5 raffle winners get the full value of their winning order back in $KITE, up to $100 each.
▷ Orders count from Nov 19, 20:00 UTC to Nov 24, 08:00 UTC. We'll announce the winners and send all rewards by Nov 28.
Get started at agentpassport.ai. Full rules: gokite.community/online/kite…
Let your agent go shopping. 🪁
Most people won't hand their card to an AI, and most shops don't take stablecoins yet. So what can an agent actually buy?
Our new Kite Agent Passport film follows a real run from Oct 13. We asked @bot for Uber credit, and it bought the gift card from @bitrefill in USDC.
▷ We asked for $10. The smallest Uber card on Bitrefill is $15, so it told us before buying anything.
▷ It created the invoice, locked the price at 15.00 USDC and paid from the $100 allowance already set in Kite Agent Passport.
▷ The card code landed right in the chat. 15.003 USDC in all: 15.00 for the card, 0.002 for the invoice and 0.001 for the code.
Your agent can shop the rest of Bitrefill the same way, from Uber rides and groceries to Airbnb stays, Steam games and phone top-ups.
Spend crypto with an agent. Keep your card. 🪁
If France cuts public spending, who breaks first? We gave an agent that one question and a 45-cent cap.
Our new Kite Agent Passport film follows a real run from Oct 27. The agent paid Sirenic for French court filings, public contracts and company accounts, plus rates, market and prediction-market data, all from its allowance, with no account or API key.
▷ It made 33 paid calls: 22 to Sirenic, the rest for rates, FX and ETF prices, prediction markets and news.
▷ Insolvencies aren't speeding up yet: 165 company proceedings in the last 30 days, against 164 in the 30 days before.
▷ Of five big public contractors, Sopra Steria is the most exposed, with 5.8% of revenue from public contracts. France's stock ETF trailed Germany's by 3.35 points over 30 days.
Confirmed cost: $0.2975, at most $0.3175 once two pending calls settle, under the $0.45 cap.
Research with an agent. Keep your change. 🪁
🆁🅸🅶🅶🅰 retweeted
You do not need a massive benchmark to evaluate a model. Five prompts are enough to start.
This AI on Air clip is from a recent episode of Lenny's Podcast. The speaker is Tulsee Doshi, Head of Product at @GeminiApp.
▷ Start small: even a mini evaluation of about 20 prompts can stress test a set of capabilities and reveal clear signal between model A and model B, without forcing you to scrutinize every piece at once.
▷ Be your own rater first: judging responses yourself, especially on long multi-step, multi-turn trajectories, forces you to define what good means; a small set of good examples gives you clarity on what you are aiming for before you scale up.
▷ Build intuition with tasks you care about: some people probe models with 3D animations and visual aesthetics, others with writing or hard financial analysis; the key is finding tasks where models already show signs of life but are not yet perfect by your definition, since that is where differences and magic show up.
What happens when your agent tries to spend more than you allowed?
agentpassport.ai has a new look, and this is our new Kite Agent Passport film for it, made from the site's own riso prints. It follows the budget demo on the new site: a task with a $25.00 budget, $8.00 already spent, and a request to buy a $42 software license.
▷ The agent tries it, and Kite checks first: $42.00 is more than the task's $25.00 budget, so it stops before paying and nothing is charged.
▷ You raise the total to $50.00 and approve it with your passkey.
▷ Kite checks the session again, and the agent pays $42.00.
Your agent can pay across 2,000+ services, strictly within the limits you set.
Kite checks before any money moves. 🪁
🆁🅸🅶🅶🅰 retweeted
感谢币安团队,这次 Kite 团队有幸登上币安亚洲学习营 Binance Campus 的舞台进行大师课分享,对我们来说也是一次难得的学习与取经之旅。
最后一天向 @sisibinance 请教了许多关于 marketing 和社区运营的实战经验,以及作为项目方如何与交易所在市场活动中深度协同配合去实现 1+1>2 的效果。
跟随龙头脚步,稳扎稳打,共建行业,让行业变得更加壮大才是正道。
月底 Kite 将作为币安区块链周的最大赞助商之一参与盛会,曼谷见!🇹🇭
@heyibinance @cz_binance @sisibinance @yayabinance @pearbinance @moonkimbinance @binance @binancezh #binanceblockchainweek #BinanceCampus #Binance
这几天一直在活动上,和很多小伙伴当面交流,聊产品、聊安全、聊这个行业接下来怎么走,收获很多。
对research干货分享直呼没听够时间太短;Jeff和大家面对面沟通讨论热烈,三天时间还是太短
做活动真的很累,每天只有遮住黑眼圈才敢出门,但看到大家很充实,也给我充满电
今天团队同学送大家返程,但刚有社区私信我,把一些和本次活动无关的事情牵扯在一起
相关当事人并未参与本次活动,这件事与币安组织的活动没有关联,请传谣者给行业留些好印象吧🙏🏻你我大家共同在这建设,这样造谣,抹黑的不只是币安,也是在污蔑我们每一位到场的每一位,在抹黑这个行业
让我们一起建设行业健康发展,一砖一瓦都留下的每个人的痕迹
最后,感谢大家对我们的关注和反馈,币安欢迎大家的批评和建议
返程的小伙伴们一路辛苦,期待BBW和大家再见👋🏻
#BinanceCampus
Any one model can write a briefing. What if your agent could pick the right model for every step, and pay each one itself?
Our new Kite Agent Passport film follows a real run from Oct 15. We asked @bot for a live Strait of Hormuz briefing, and it paid for five models from four labs through @BlockRunAI, with no account or API key at any of them.
▷ Before spending anything, it priced every paid call: about $0.78 in total, with no single call over the $0.50 cap we set.
▷ Claude wrote the analysis, Grok argued against it and Gemini fact-checked it. The analyst said Critical, the red team pushed back, and the final call was High.
▷ Two image models drew the infographic. Eleven paid calls in all, $0.71, paid from the $100 allowance already set in Kite Agent Passport.
Research with an agent. Keep your keys. 🪁
GDAY KITERS 🪁
One important part of building an agent economy is figuring out how AI agents will handle money.
Think about it.
An AI agent may need to purchase data, access paid APIs, use computing resources, or pay another service to complete a task.
But what kind of money should it use?
This is where stablecoins become important.
@GoKiteAI uses stablecoin-based payments to help agents interact with digital services without relying entirely on traditional payment systems.
Here's how it works:
1. Stablecoin payments
Stablecoins such as USDC provide a dollar-denominated payment option for agents. This makes it easier to price services and process transactions without the same price fluctuations associated with volatile crypto assets.
2. Automated payments
Through Kite Agent Passport, users can authorize spending sessions with specific budgets and permissions. Agents can then make eligible payments within those approved boundaries.
3. Cross-chain support
Kite is working toward making agent payments possible across multiple blockchain networks. This matters because agents shouldn't be restricted to services operating on just one network.
4. USDG support
Kite has also introduced support for holding and sending USDG on Robinhood Chain, including gas-sponsored transfers and automatic USDC-to-USDG routing.
This is particularly interesting because it reduces some of the friction involved in moving between supported stablecoin payment options.
Why does all this matter?
The agent economy will require more than intelligent software.
Agents need a practical way to pay for the services they use, receive value, and interact with other applications.
Stablecoins provide a useful payment medium, while Kite provides infrastructure designed to connect those payments to AI agent workflows.
The bigger picture is simple:
As AI agents become more independent, stablecoins could become one of the financial tools powering their everyday activities.
Kite is building toward that future.
HIGHER 🪁
@Kite_Frens_Eco @0xLaughing
🆁🅸🅶🅶🅰 retweeted
Creating one agent for each human role, a data scientist agent, an engineer agent, a PM agent, may rest on a flawed assumption from the start.
This AI on Air clip comes from a recent episode of The Cognitive Revolution Podcast. The speaker is Flo Crivello, founder and CEO of Lindy. He argues that in most cases you should consolidate as much as possible into a single agent rather than splitting work across many.
▷ Human organizations divide labor because each person has only 24 hours a day and can hold only so much context. Agents have none of those constraints: they can fork, duplicate, and do as much as they want all day.
▷ So division of labor is not a good reason to have multiple agents, and people often over-anthropomorphize, instinctively building systems with far more agents than optimal.
▷ He adds that this is not absolute: there are still valid reasons for multi-agent setups, but division of labor is not one of them.
🆁🅸🅶🅶🅰 retweeted
Kite fam, see you in the Space!
Join @Henryleemr, @open_founder and @reppo to talk agentic payments, AI data markets and verifiable agents.
Today, Nov 4 at 3 PM UTC / 11 PM SGT.
Set a reminder and tune in. 🪁
We’re entering a new paradigm where AI agents make decisions and take actions that demand reliable reasoning.
Tomorrow at 3 PM UTC @open_founder hosts @reppo and @GoKiteAI on Spaces to talk agentic payments, AI data markets, and verifiable agents.
Set your reminder below.
GN KITERS 🪁
One interesting question about the future of AI is this:
What happens when AI agents become active participants in the digital economy?
Think beyond agents answering questions or automating simple tasks.
Imagine an agent that can access financial services, interact with DeFi applications, move digital assets, and execute transactions across different networks to complete a task.
For this to work, agents need more than a wallet. They need identity, authorization, and infrastructure that allows them to move value securely.
This is the direction @GoKiteAI is exploring with what it calls Agentic Value Movement.
Through Kite Agent Passport, users can define what their agents are allowed to do, which services they can access, and the conditions under which they can transact.
The bigger idea is to connect AI agents to the existing digital asset ecosystem rather than build an entirely separate financial system for them.
Why does this matter?
Because as agents become more capable, they will need access to the same financial tools humans and businesses already use.
Kite wants to help make that possible across applications, assets, and blockchain networks.
We're moving toward an economy where AI agents won't just process information. They'll also participate in moving value.
And that's a development worth watching.
The next evolution of DeFi might not just be more users. It could be AI agents becoming participants in the ecosystem.
@Kite_Frens_Eco @0xLaughing
Meet your Muse. Meet the 16 winners. 🪁
30,095 submissions. 16 Muses drawn.
1. @_rumaisah
2. @tolgaozek
3. @wolfngel
4. @1cewithmeenaeth
5. @tradeguru
6. @shahariar1k
7 to 16: @kerrylegit @KingTowdah @Jok3rjo6 @oladipsingami @apaajadesainin @bestan777 @Elowen_77 @QNam30992 @BabuLaki47370 @0xAlphana
1,000 USDC + $1,100 in Kite Router vouchers, split by rank.
Winners: a DM from @GoKiteAI is on its way for the account checks. Prizes go out once the checks clear.
Thank you to everyone who created, posted and entered. Your Muses made our feed.
Your photo. Your Muse. 🪁
Create your @Muse-style PFP for free with Kite Agent Passport. Share it on X for a chance to win.
🎁 16 winners.
1,000 USDC + $1,100 in Kite Router vouchers.
Create. Post. Enter.
Here’s how ⬇️
From more efficient computer-use models and open weights to multi-model agent platforms and frontier safety, the period since our last edition brought major developments across the AI-agent stack.
Here’s everything worth catching up on in the latest edition of AI Agents Pulse:
1️⃣ OpenAI released GPT-6.1 Terra for visual reasoning, computer use and writing in ChatGPT Work, Codex and the API. Standard API pricing is $1 per million input tokens and $5 per million output tokens.
openai.com/index/introducing…
2️⃣ Google announced Gemini 4 Carbon for coding, cybersecurity and long-horizon tasks. Access is starting with a subset of cyber defenders through the Fairwind Program, ahead of broader developer, business and consumer availability.
blog.google/innovation-and-a…
3️⃣ Mistral released improved Mistral Large 4 weights for agentic coding and defensive cybersecurity, with Low, Medium and High reasoning modes. Its modified MIT license requires a separate commercial license for businesses operating a Model-as-a-Service offering.
mistral.ai/news/mistral-larg…
4️⃣ Liquid AI released the first LFM3 models in the 2B-parameter range, bringing text, vision, audio, omni and decision models onto a shared backbone. Base and post-trained weights are available on Hugging Face, alongside Playground and API access.
liquid.ai/blog/lfm3
5️⃣ Google announced Genie 4, a world model for real-time interactive simulation with native audio. Access began with limited trusted testers; Project Genie access for Google AI Ultra subscribers and Gemini API early access are still coming.
blog.google/innovation-and-a…
6️⃣ GitHub made Agentic Workflows generally available. Developers define automation in Markdown, which compiles to GitHub Actions YAML, using Copilot or another coding agent. Agents run with read-only permissions by default, with proposed changes checked before application.
github.blog/changelog/2026-1…
7️⃣ Microsoft introduced the Copilot harness in Foundry Agent Service in public preview. It manages the agent loop and sandbox for long-running tasks, while developers supply instructions, skills, tools and model choice.
devblogs.microsoft.com/found…
8️⃣ GitHub introduced the Copilot app and Copilot Fusion at Universe. The app connects agents and tools across projects, while Fusion chooses how to execute a task across one or more models.
github.blog/news-insights/co…
9️⃣ OpenAI released GPT-6 Rosalind for multistep life-sciences research with scientific tool use. Access is limited to organizations approved for Rosalind Research and their authorized users, rather than general API availability.
developers.openai.com/api/do…
🔟 OpenAI updated its Model Spec with guidance on context, oversight and human agency, organized around five values. The company said its models do not yet fully reflect the intended behavior described in the Spec.
openai.com/index/updated-mod…
1️⃣1️⃣ OpenAI detailed expanded monitoring for frontier agents across training, evaluation and internal use, with adaptations for external deployment. Its layered system combines activation probes, model reviewers, investigative agents and human response to flag unauthorized or deceptive behavior.
openai.com/index/how-we-dete…
Most people won't hand their card to an AI, and most shops don't take stablecoins yet. So what can an agent actually buy?
Our new Kite Agent Passport film follows a real run from Oct 13. We asked @bot for Uber credit, and it bought the gift card from @bitrefill in USDC.
▷ We asked for $10. The smallest Uber card on Bitrefill is $15, so it told us before buying anything.
▷ It created the invoice, locked the price at 15.00 USDC and paid from the $100 allowance already set in Kite Agent Passport.
▷ The card code landed right in the chat. 15.003 USDC in all: 15.00 for the card, 0.002 for the invoice and 0.001 for the code.
Your agent can shop the rest of Bitrefill the same way, from Uber rides and groceries to Airbnb stays, Steam games and phone top-ups.
Spend crypto with an agent. Keep your card. 🪁
GM KITERS 🪁
AI agents are getting smarter, but intelligence alone isn't enough.
Imagine an AI agent that can research markets, pay for data, access different services, and complete tasks on your behalf without you having to manage every step manually.
Sounds like the future, right?
But for this to work, agents need more than intelligence. They need identity, payment infrastructure, and clear permissions that determine what they can do and how much they can spend.
This is where @GoKiteAI comes in.
Through Kite Agent Passport and its agent-focused infrastructure, Kite is working toward a future where AI agents can pay for digital services, interact with other systems, and execute tasks within limits defined by their users.
Think about businesses using agents to gather information, analyze data, and access paid tools automatically. Instead of humans handling every transaction, agents could manage eligible payments as part of their workflow.
That's what makes the agent economy interesting.
We're moving from AI that simply answers questions to AI that can actually get things done.
Kite is building toward that future, and I'm interested to see how far this ecosystem can go as adoption grows.
What industry do you think will benefit the most from the agent economy: finance, e-commerce, or digital services?
#KiteBBW
Binace square post link: app.binance.com/uni-qr/cpos/…
1
2
34
GM KITERS 🪁🪁
For a while now, I haven’t been very active here because school has been keeping me busy.
But while I was away, @GoKiteAI kept building, and there’s quite a lot to catch up on.
So today, let’s talk about Kite AI, its recent developments, and why I think the direction it’s taking is worth paying attention to.
First, what problem is Kite trying to solve?
Think about how we use AI today.
You can ask an AI to research a product, analyze data, write code, or plan a business campaign. But what happens when that AI needs to pay for a premium API, purchase a service, or complete a transaction to finish the task?
That’s where things get complicated.
AI agents need more than intelligence to perform useful work independently. They need an identity, clear permissions, access to payments, and a way to prove what they have done.
This is the problem Kite AI is working to solve.
Its goal is to build infrastructure that allows AI agents to transact and coordinate while remaining within boundaries defined by their users.
And some of its latest developments show how this vision is taking shape.
1. Kite Agent Passport is becoming more practical
One of the most important parts of the ecosystem is Kite Agent Passport.
It allows an AI agent to make payments on behalf of a user without giving it unrestricted control over the user's funds.
Instead, the user can approve a spending session with specific limits, such as a budget, an expiry time, and the services or assets permitted.
Imagine asking an AI agent to research a topic using paid services.
You could authorize a specific budget for that task. The agent can then make eligible payments within the approved limits, while its activity remains available for review.
The important distinction is that the agent receives controlled authority, not unlimited freedom.
That’s a meaningful step toward making AI agents useful beyond conversations.
2. The Passport dashboard and service discovery have improved
Kite’s October 9 update introduced a redesigned Passport dashboard, support for Arc mainnet in the wallet, and a public Explore page where people can browse services and recipes without signing in.
Why does this matter?
Because infrastructure is only useful when people can understand and use it.
A developer building an AI agent needs to discover available services. A user needs to understand payment activity and spending sessions. And both need a clearer way to interact with the system.
Kite is working on these practical parts of the experience, not just the big idea of an agent economy.
The same update also introduced an Aave auto-repay feature for eligible USDC loans on Base, allowing users to set a health-factor threshold and approve the action in advance.
It’s an example of how agent infrastructure could support specific financial workflows rather than simply answering questions.
3. Kite is exploring more advanced agent payments
Another important development is the continued work around x402 payments.
In simple terms, x402 enables a service to request payment when an agent tries to access paid content or an API. The agent can then handle the payment process and receive the service response.
Think of an AI research agent that needs information from several paid data providers.
Instead of relying on a traditional subscription for every service, the agent could pay for eligible resources as needed, subject to the permissions and payment rules in place.
This opens the door to a different way of delivering digital services: machine-to-machine payments, where software can pay other software to get work done.
Kite Passport has also been developing multi-offer payment flows, helping agents handle situations where a service offers different payment options.
The bigger idea is to make payments a natural part of an agent’s workflow.
4. Cross-chain payments and network support are expanding
@Kite_Frens_Eco @0xLaughing
1
1
51
Kite is now live on @Arc Mainnet, building the foundation for an agent-to-agent platform with trust, governance and USDC payments.
The future of the agentic economy is here.
Arc Mainnet: arc.io/
Kite Agent Passport: agentpassport.ai/
Arc Mainnet is live.
Arc launches as the Economic OS for the internet: an open platform for global markets, real-time value movement, tokenized assets, and agentic economic activity.
Arc is more than a blockchain.
It launches as a full-stack financial platform with assets, applications, interoperability, developer infrastructure, and Circle platform services live from day one.
Arc delivers USDC as native gas, deterministic sub-second finality, EVM compatibility, and institutional validators.
It integrates with Arc Studio, App Kits, Arc Portal, Circle Agent Stack, CCTP, Gateway, CPN, and StableFX.
A complete economic platform at genesis.
Arc launches with infrastructure for:
→ Agentic economic workflows
→ Lending and borrowing
→ Trading and liquidity
→ Onchain FX
→ Payments and settlement
→ Tokenized assets
→ Exchanges, wallets, custody, compliance, data, and developer tooling
190+ institutional and ecosystem builders are building across Arc.
🆁🅸🅶🅶🅰 retweeted
在办公室听 @scottshics 讲产品有一种回到学校的错觉
好的团队是静得下来的,脚踏实地去 build
好事多磨,静待佳音
2026 may be the Kite year 👀 @GoKiteAI
Whiteboarding with @GoKiteAI team last three weeks in person. Something big is coming soon
🆁🅸🅶🅶🅰 retweeted
Kite Agent Passport 2.0
iykyk 👀
Agents should not discover their authority halfway through a negotiation.
It should be set before the negotiation begins.
Kite Agent Passport 2.0 shows what governed agentic commerce looks like in procurement:
▷ A person defines the target price, the limit, and the terms that stay protected.
▷ Both agents negotiate within their own private mandates, with every offer checked before it is sent.
▷ Only a package permitted by both mandates reaches human approval and signature.
Kite Agent Passport does not replace the human signer. It makes clear what an agent can and cannot do before it acts.
Watch the procurement demo. 🪁



























