Tony0x.eth

Ledger office conference. With #Figment and #Solana x.com/i/broadcasts/1…

Introducing my first vibe coding project: ClawFeed--Curing Info Anxiety with AI. I follow 5,000 people on Twitter. Spent 2 hours a day scrolling. Still missed what mattered. So I built an Agent to read everyone's tweets for me. Now it takes 5 minutes. Meet ClawFeed: clawfeed.kevinhe.io github.com/kevinho/clawfe… I tried RSS, Pocket, newsletters — All just moved information from one inbox to another. The real problem isn't "too much info." It's "filtering costs too much." Every "is this worth clicking?" decision burns attention. ClawFeed: AI reads everything. You read what matters. Every 4 hours → structured AI brief Daily → 5,000 feeds compressed into ~20 highlights Mark anything → instant AI deep dive Format: @username + their actual words. Not "industry discusses" — "@karpathy says transformers aren't the endgame." 10 days of real data: Before → After: 2 hrs scrolling → 5 min reading briefs Saved 500 articles, read 5 → mark it, AI analyzes instantly 100% noise exposure → AI filters 95% Always FOMO → important stuff surfaces automatically 54 structured summaries. Zero missed signals. How it was built: Zero framework dependencies — Node.js native HTTP + better-sqlite3. Under 50MB memory. From v0 (Telegram + Markdown) → v0.5 (OAuth multi-user Dashboard) in 4 iterations. Published as both an OpenClaw Skill + Zylos Component. One command to install. Open source: github.com/kevinho/clawfe… MIT license. Clone, install, run. Hosted: clawfeed.kevinhe.io No signup needed. Just open and go. Read less. Know more. Built by @OpenClaw & @ZylosAI

OpenAI drops GPT-5.4 mini & nano — smaller, faster versions of their flagship model optimized for coding, multimodal tasks, and high-volume API workloads. Developers, this could be your new go-to for speed at scale. 🚀 #AI #OpenAI

A global virtual hackathon for Opus 4.6 🌐 13,000+ applicants. 500 selected. 227 projects. 6 finalists. 3 winners. Final presentations in front of @bcherny & @_catwu Attendees flying in to be at @SHACK15sf Safe to say, Claude Code’s 1st birthday was one for the ages ⚡️

An unknown entity holding $30M USDC across two Hyperliquid accounts is currently short 330.27K CL ($32.5M) and 144.71K BRENT ($15.8M), making him the largest oil short onchain. Out of the $577M total OI across both markets, he controls 8.37% ($48.3M), all on the short side. He is currently down $3.3M with a average entry around ~$109.2 on BRENT and ~$91.4 on CL. hypurrscan.io/address/0x985f… hypurrscan.io/address/0xac03…

Running crypto AI agents just got 200x more efficient & powerful 🔥 Introducing the CoinGecko CLI — a free, open-source terminal for real-time crypto market data, built for AI agents (such as @openclaw, Claude Code) & developers. #BuildwithCoinGecko See how it works 👇🧵

🚨NEW: In approximately 2-3 hours the 20,000,000th bitcoin will be mined. This is a historic day. 🔥

"Bitcoin is way too transparent" - @RayDalio Here's how the Dutch government moved €15 BILLION in Gold early 2023 (very obscure)👇

The Korean government seized a ledger, and uploaded the seed phrase to the photo Out of 2,333 bitcoin seized they have lost 1,742 of it 😭

lesson from stablecoins: the primary tech is redemption, not “stability.” pegs are governance theater until there is a deterministic path back to underlying value (think why we have peg stability module.. its clear redemption). the systems that survive are the ones that make redemption constraints explicit, define who gets served first, and accept that scarcity has a price. stability is a function of credible convertibility under stress, plus a balance sheet that does not collapse when everyone tests the off ramp at once. lesson from mev: if value can be extracted, it will be extracted. the only question is whether it is latent and adversarial or explicit and auctioned. mev taught the ecosystem to stop moralizing and start engineering. once you accept extraction as a physical law, you build mechanisms that surface it, price it, and route it to the right place. that shift is why auctions beat backroom deals, and why transparent ordering rules matter more than narratives. the meta lesson: markets scale when exits are explicit, ordering is deterministic, and extraction is priced (auctioned, liquidated, cleared). once those are true, you are not selling a story, you are selling an instrument, and instruments compose into more instruments. liquidity needs to be on a state machine because liquidity is just displaced time. why do you think we call it time value of money even on the perblock level (scarcity, risk, optionality). discretion obfuscicates neutrality. this is what @VitalikButerin has alluded to with having bias on your chain. it slows things down the more neutral the surface, the lower the effective coefficient of friction, pushing the velocity of money toward its theoretical limit there is a reason we went out was able to engineer reinsurance around the sUSDai instrument, even with a tradfi partner (munichre). you rarely see that reflex in “private credit” because most private credit is a closed wrapper around relationship risk, marked smooth, redeemed through windows, and defended by discretion. it is not designed to clear in public under stress, so it is not designed like an insurable product, which is why it wasnt really invented until we shared our model (removing idiosyncratic risk so its purely against the chip and the depreciation of the chip) @USDai_Official is not really private credit. it is collateralized borrow against chips with explicit settlement rules. the risk is defined, the exits are mechanized, and the market can price duration and urgency continuously. once you wire credit that way, an interest rate can actually form. for something that seemingly depreciates too fast or rather something that pays you back too fast for tradfi's tastes, this is how you scale the most important market in the world today: compute

Protocol-level privacy on existing blockchains vs private blockchains for stablecoin flows. There are a few ways to achieve privacy on blockchains. First, there is protocol-level privacy. It means privacy is achieved on top of existing public blockchains. Second, there is blockchain-native privacy. Self-explanatory. By privacy, I mean confidentiality of balances and flows, and unlinkability between sender and recipient. In payments, privacy is not just a feature, it’s a supply chain: it must persist through settlement, treasury, and payouts. Now let’s apply privacy to the major use case we have today: payments. The majority of stablecoin payments are happening on three chains: Ethereum, Tron, and Solana. This matters because these chains already have liquidity, wallet distribution, exchange support, and merchant adoption. Let’s imagine a PSP is settling privately for a merchant on Tron. The merchant uses privacy via Hinkal and receives USDT on their current wallet. The merchant needs privacy not only to make payouts to vendors and payroll. If payouts happen on the same chain, the same thing happens: the payout is private. The recipient can receive money privately and then make another payout without revealing what they have in their wallet. This cycle ends only at the leg where fiat is needed. On the fiat leg, the end recipient becomes public and off-ramps. In this case, the economy is circular: it stays in stablecoins and allows payments between everyone. It happens on chains everyone already trusts, with stablecoins everyone already trusts, and with wallets everyone already trusts. Privacy becomes a native part of the stablecoin economy without forcing any migration. Now let’s imagine that settlement happens on a private blockchain. Let’s imagine the name of this private blockchain is Corona (or any new privacy-focused chain). Corona makes it possible to onboard the merchant onto their blockchain, but with limited wallet support and a wrapped version of the major stablecoin USDT. The merchant is now on the Corona blockchain, but what is the probability that the vendors are also on the same blockchain? Not high, since they need to use the same wallet and there needs to be liquidity on the chain. More importantly, vendors have no incentive to adopt a new chain just because one merchant wants privacy. So the merchant has to use a bridge to make payouts on major chains. And what makes it even more problematic is that the counterparty has no way to receive confidential settlement. So it becomes public how much they received, and the privacy supply chain is broken. I think in the future we will see infrastructure that tries to marry these worlds, so it can send from the Corona chain into the recipient’s shielded account on a major chain. But public chains aren’t sleeping either, and they are aggressively competing for payments adoption. Even with perfect bridges, distribution and wallet support don’t migrate overnight. With protocol-level privacy deployed on major chains, privacy can be implemented without intermediating any of the flows: chain, stablecoin, or wallet. Chain-level privacy brings fragmentation of the payment supply chain. It creates a closed island where privacy exists internally, but breaks the moment value leaves the island. Protocol-level privacy can also work cross-chain. For example, USDT from Chain A can be moved to Chain B and still remain inside a shielded pool. Because shielded pools can exist across chains, the user experience can stay consistent even while liquidity moves. With chain-level privacy, this is much harder, because the privacy environment is isolated to a specific chain. Stablecoins win because they are universal money, and payments need universality more than they need new ecosystems. Privacy should upgrade the rails the world already uses, not force the world to move onto new rails.

Stop prototyping. Start shipping. The inaugural SIDUS AI Hackathon is here. We're challenging you to build powerful AI agents and analytical tools on a groundbreaking platform. 🎯 Compete for a $12,000 prize pool and get your work seen by the best in AI. Your next big project starts now. 👉Register & Details: hackathon.sidusai.ai

1/ Meet the threat actor John (Lick), who was caught flexing $23M in a wallet address directly tied to $90M+ in suspected thefts from the US Government in 2024 and multiple other unidentified victims from Nov 2025 to Dec 2025.

1/ Crypto, tokenization and stablecoin were prominently featured at @wef Davos this year. Institutions are still learning but there is a growing understanding as well as conviction that the whole digital assets ecosystem is important and valuable to the world. Some crypto related highlights 👇

I'm solo founder competing against 10-person sales teams My secret weapon: an AI that qualifies leads in 42 seconds See how it works 👇

CRM and the holiday outreach did their job, but the day still turned out unexpectedly intense 😅 — the agent pitched a non-existent trading bot, costs suddenly spiked, and Telegram blocked our domain. 1️⃣ Conversation review — new edge case: the agent actually booked a demo with a lead by pitching a solution… that we don’t even have yet 🤷‍♂️ 2️⃣ Finished the Christmas outreach 🎄 — huge wave of replies → agent got overloaded → spend increased 💳 3️⃣ Telegram stopped sending webhook events to n8n just because it didn’t like our domain — we switched the domain and everything started working again ⚙️ If you want to pull all leads into one CRM and never lose a conversation again — DM us 💬

Although we’re still tweaking the CRM, the main focus right now is on “raising” our Zero Inbound agent — teaching it to behave like a polite sales rep, not an annoying spammer 🤝📩 Yesterday we kept polishing the Zero Inbound agent’s behavior in real conversations and baked a set of important changes into the prompt 🛠 1. Asking about budget — without an interrogation 💸 Previously, the agent could circle back to the budget question several times in a row and came off as pushy. Now the prompt explicitly says: • the agent asks about budget only once before suggesting a call, for example: “By the way, do you have a budget range in mind for tools like this?” • if the lead doesn’t answer or changes the topic, the agent doesn’t push and never asks about budget again 🚫 2. No Calendly — only @manizha_business 📆➡️👩‍💼 The agent sometimes tried to be proactive with Calendly and attempted to book calls directly in chat. Now: • it never suggests Calendly and never books calls itself • if a lead asks why there’s no Calendly link, the agent explains that the calendar is overbooked • for anything call-related, the agent always redirects to @manizha_business 3. The correct email ✉️ Before, the agent was giving out a made-up email: tony@palo-alto.ai Now the correct address is hard-coded into the prompt: partnerships@palo-alto.ai 4. Budget — only if we’re talking about the CRM 📊 Sometimes the agent asked about budget completely “out of the blue”, even when the CRM hadn’t been mentioned yet. We’ve tightened that up: • the agent never asks about budget if the chat hasn’t mentioned the CRM (our product) before 5. No promises to pay for anything 💳🚫 Previously, the agent could casually say we were ready to pay for someone’s services. Now there’s a strict guardrail in the prompt: • whenever payment or money comes up, the agent clearly says it’s an AI agent and can’t handle payments • for any billing or financial questions, it redirects to @manizha_business In parallel: the holiday broadcast continues 🎄📨 The holiday messages are still going out, and along the way we found another important CRM-related detail: • right now, if a user’s account is deleted, we only see an error when trying to message them ⚠️ • but the fact that the account is deactivated isn’t stored anywhere explicitly in the database The next step is to start marking deactivated users directly in the CRM ✅ So the system understands that this contact is no longer reachable and doesn’t try to send them anything in future campaigns 🔁 If you need a CRM tuned for your lead flow, we can build it around your use case 💬📈

Yesterday we put our CRM and bulk messaging to the test in the real world — sending holiday greetings from all our accounts at once 🎄📨 🎯 The task We’ve got dozens of accounts, and the relevant leads we want to congratulate are scattered across all of them 🌍 At the same time, our current broadcast system can’t “fire” from any account we choose — each campaign is tightly tied to a single account and its chats 📲 🧩 How we handled it Instead of one “magic” blast, we built a cascade: • we moved from smaller accounts to bigger ones 📈 • had to create dozens of separate campaigns 📑 • our AI agent inside the CRM helped a lot — it sped up preparing and configuring all these broadcasts 🤖 Then we launched the holiday campaign and are now actively monitoring how it’s performing 👀📊 🚧 Telegram limits & manual tweaking Of course, it doesn’t work perfectly in full auto — we hit Telegram’s limits: • even in “warm” chats, some accounts can’t message users if the last conversation was too long ago 🕰 • because of that, part of the campaign has to be pushed manually — we track who actually got the message and who didn’t ✅❌ Still, the large holiday blast across dozens of accounts went live 🚀 In parallel, we’re collecting insights on how to make this format more automated next time ⚙️ If you want a CRM that actually works, reach out 💬

#Biz_Achieves_Diary — 25/12/’25 How are sales going? I’ve been testing a new hypothesis: a full CRM subscription should be cheaper, because founders without budget simply won’t be able to buy it and outreach investors. But even at $200, selling a subscription is not as easy as I expected. By the way, about yesterday’s interesting calls: 1\ Peeijay Founder of Web3 Influencers Network Building their own project — Norah Currently raising for the project, offering marketing support 2\ Andrey Sintetik AI (silver.sintetik.ai) Beta is live, users onboarded Looking for advisors and partners =============== Intro groups from yesterday: • Songjam 🤝 Iron Key Capital • Pat 🤝 aZen VC • Pat 🤝 Poolify Lab * 25+ more intro groups after just 10 calls… If you also want to automate leadgen… message me. We’ll build it for you.

#Biz_Achieves_Diary — 24/12/’25 turns out cutting down the prompt isn’t as easy as I thought you remove one thing — and the agent already stops following the script… still doing fine-tuning of the agent with the team speaking of interesting calls from yesterday: 1\ Paras • support projects with KOLs • strong network of KOLs, exchanges, launchpads • marketing for early stage 2\ Tommy | DexTools • launchpad + trading platform • doing IDOs for different AI projects • now building a platform around AI agents =============== intros from yesterday: • Songjam 🤝 LDA • Songjam 🤝 SevenX Ventures • Pat 🤝 aZen VC and 25+ more intro groups after 10 calls… If you also want to automate lead gen and calls with VCs, launchpads, etc. — DM me, we’ll set it up.