Lost in Web2, found in Web3! Building freedom through decentralization!

Joined December 2017
Marques Ken retweeted
AI trading agents are getting better, but how do they prove they can trade well enough to access capital? That’s the problem @agenticscredit is working on. The project is building a credit system for human traders and AI agents, using their actual trading history to assess performance and risk. The main part is the Agentic Credit Score (ACS), which runs from 300 to 850. It looks at factors like profitability, drawdowns, consistency, win rate and trading history, with greater weight given to real-money results than paper trading. Here’s where it gets practical 👇 A score of 580 or above meets the current qualification threshold for the funding waitlist. Below that, traders can continue building their track record through paper trading. The platform also has an Agent Builder for creating and adjusting trading strategies, plus a Risk Engine that enforces limits on losses, leverage and position sizes. So the process connects three things: building an agent, proving its performance and potentially qualifying for trading capital. Funding still depends on allocation and the applicable terms, but the idea of using measurable trading performance to assess credit is worth paying attention to. Agentics Credit has also launched its Quest Hub, where users can explore available quests and compete for a share of $5,000,000 in rewards. If you're building AI agents or interested in how trading performance could become a form of credit history, this is worth exploring. Check out the project and its Quest Hub here 👇 agentics.credit/ xitter.cf/agenticscredit/status/…
Your trading agent can already execute. Proving it deserves capital is the hard part. That's where Agentics Credit comes in: the credit layer for autonomous trading. Connect and start paper trading today to win your share of $5,000,000. agentics.credit/launch
89
11
146
4,342
Marques Ken retweeted
Crypto AI is entering a new phase, and the next winners may not be the tokens getting the most attention. September was a strong month for Grayscale’s AI Crypto Sector, which gained roughly 54%, outperforming the broader crypto market. But recent weakness is a reminder that AI-related tokens still face valuation concerns, especially as questions about the AI bubble grow. What interests me most is where development is heading. Three areas are worth watching: 1. AI agents and payments Agents need a way to pay for APIs, data, computing power and model access without relying on traditional accounts or subscriptions. Coinbase’s x402 protocol is helping explore that model through crypto micropayments. Chainalysis has reported more than 138 million payments on @base, highlighting how stablecoins could support machine-to-machine commerce. 2. Decentralized AI infrastructure Projects are targeting real needs across the AI industry: licensed training data, GPU financing and computing resources. @bittensor is building decentralized machine learning markets, @NEARProtocol is exploring agentic applications and private computing while projects like @prlnet are experimenting with turning AI-related computation into useful work for blockchain networks. The important question is whether these systems can deliver measurable value at a competitive cost. 3. Real usage over token speculation The previous wave of AI agent tokens showed how quickly a promising narrative can attract capital, and how sharply valuations can fall when attention fades. The next phase needs stronger evidence: paying users, sustainable revenue, verifiable workloads and products that solve problems better than existing alternatives. There are still major obstacles, including security risks, centralized control of computing resources regulatory restrictions and the difficulty of matching centralized AI systems on cost and performance. the long-term opportunity may not be in tokens that merely carry an AI label but in the infrastructure that makes autonomous agents useful. Crypto AI is still early, and not every project will survive.
42
5
78
440
3,035 stock and ETF-linked rToken assets That is the number Bitget has reached just as markets are heading into one of the more interesting FOMC meetings of the year 👀 The rate decision itself may not be the biggest surprise. The Fed raised rates to 3.75%–4.00% in September, and markets are now heavily leaning toward a hold. What matters more tonight is what the Fed says about the next step. The backdrop is messy Oil is still trading at unusually high levels after the latest Middle East escalation, keeping energy inflation in the conversation. The 10-year Treasury yield is hovering around 5%, the dollar is near a two-week high, and the Fed is trying to judge how much more tightening the economy can absorb without creating unnecessary damage At the same time, markets are not moving as one block Gold has pushed back above $4,000 even with yields still high. Bitcoin has been relatively quiet around $80K. U.S. equities rallied into the meeting, with tech once again leading as the market tries to balance strong AI spending against increasingly expensive capital. That divergence is what I find more useful to watch than the rate headline itself. If the Fed stays on hold but keeps December open, I want to see whether yields and the dollar move first, and then how that filters into growth stocks, gold, oil and crypto. If the tone is softer than expected, the transmission could look completely different. This is also where Bitget reaching 3,035 stock and ETF-linked rToken assets becomes interesting in context. Instead of watching FOMC only through $BTC or ethereum:native , the same USDT-based environment now gives access to supported markets linked to names such as $AAPL and $NVDA , broad ETFs like SPY and QQQ, and newer additions including Tencent, BYD and Xiaomi. For me, the value is not having 3,035 things to trade. It is having a much wider market map when one policy decision can move rates, currencies, commodities, equities and crypto in different directions. On Bitget, supported rToken/USDT pairs can be found under Stocks, where price, spread and available order types can be checked before making any decision. rTokens provide exposure linked to the underlying stock or ETF and are not the same as directly owning registered shares through a traditional broker. I’m keeping this exact cross-asset watchlist for after the FOMC. Once the decision and press conference are out, the more interesting question will be simple: what did the market actually reprice? Read more here: bitget.com/support/articles/… This content for informational and educational purposes only. Not investment or financial advice, and not an offer or solicitation to trade. rTokens and digital assets involve market, liquidity and product-specific risks. Availability and eligibility may vary by jurisdiction
120
19
1
109
2,589
Marques Ken retweeted
most memecoin launchpads reward whoever gets in earliest @zerufinance is exploring a different model with activityfi instead of starting with a token and hoping a real community forms around it, zeru starts with an economy that already exists trading activity becomes the basis for an activity market, and the people contributing to that economy can earn the first seats in the resulting ztoken that changes the question from who got the earliest insider allocation to who actually helped build the underlying activity i think this is the more interesting direction for memecoin launches, especially when onchain behavior can be measured rather than judged by follower counts alone activityfi is worth watching for anyone interested in how tokens could emerge from real onchain communities check your wallet history and explore the new model 👇 app.zaps.markets/?invite=Dct…
224
28
200
10,863
Marques Ken retweeted
building in crypto often comes with a trade-off. you might prefer Solana for its fast execution and low transaction costs but a lot of developers are already working with Solidity and the EVM. meanwhile, users and liquidity are spread across different ecosystems. so what if developers didn't have to treat them as separate environments? that's an interesting problem @RomeProtocol is working on. here's how it works: → Rome runs EVM inside Solana as a single Solana program → Solidity contracts can call Solana programs within the same atomic transaction, without an extra bridge hop → EVM users can sign with their usual wallets, while Solana users use their own Solana keys → both sides can interact with the same contract, state and liquidity the part worth paying attention to is how these pieces work together, rather than simply moving assets from one chain to another. Rome EVM has been live on Solana mainnet since June with an open-source implementation audited by Halborn. it'll be interesting to see what developers build when they can combine Solidity and Solana programs without splitting the experience into separate environments.
135
16
148
3,396
Marques Ken retweeted
ok so easiest way i've found to explain this your apartment building uses the same lock design on every door. someone finds a way to break it. now what? you can't just fix your door. everyone waits for the building to redo all of them @Starknet accounts skip that problem. each account is basically a tiny program that picks its own lock so if a better lock shows up, an account can swap it. no waiting on the whole network and it's not just a whiteboard idea. an account with a lock built to resist quantum computers was already demoed on Starknet mainnet (Quantum Hub, Aug 2026) still a demo tho, and old accounts won't switch on their own that's what cryptographic agility means to me. not "we're safe forever", just "we can change the locks" which raises the part i keep going back and forth on: once these locks are ready, should wallets swap them for users by default, or should it always be your call? ngl i lowkey can't pick a side yet quantum.starkware.co/
96
11
184
1,735
Marques Ken retweeted
Meme launches move fast, but getting in early shouldn’t just come down to who clicks the fastest. That’s where reputation can change the way onchain launches work. At The Bridge by TheTieIO, @zerufinance is bringing Activity Markets into the spotlight alongside trust oracles and the future of onchain markets. → ZeruAI analyzes wallet behavior to understand how users actually participate onchain. → zScore turns that activity into a reputation signal, giving past behavior a role in determining future access. → Activity Markets are coming soon to Zaps, featuring community-driven launches, no dev bag, and early access based on previous trading activity or clan participation. The idea is simple: your onchain history shouldn’t mean nothing when a new opportunity comes around. If you’ve been active onchain, now’s a good time to connect your existing wallets and check your starting Zaps balance. Because when the next launch happens, should early access go to the fastest clicker or the people who have already contributed to the ecosystem?
137
13
122
1,866
Marques Ken retweeted
Most crypto projects promise you yield, but can they explain where the yield comes from? @Oly_Dao has 3 sources of rewards: >ETH staking through Lido >Fees earned from providing liquidity on Uniswap >Trading taxes collected from its own pools Now robinhood:0x8e8266b11eaff21998b31c5ad258198e103e1fb0 is bringing this model to Robinhood Chain. It provides liquidity for tokens linked to NVIDIA and Tesla stocks, as well as BTC and ETH. The goal? Become one of the biggest liquidity providers on Robinhood Chain. Why does this matter? These stock tokens need enough buyers and sellers to trade easily. That means liquidity matters. So, next time a crypto project promises you yield, ask one simple question: Where is the yeild coming from ?
70
28
1
151
5,369
If I bought an inscription and couldn’t see it in MetaMask, I’d want an explanation before doing anything else. That’s a useful place to start with @beepzfun. Its beepz-20 inscriptions aren’t ERC-20 tokens. Balances exist inside the Beepz contract on Robinhood Chain, alongside listings and inscribe rules. The site, or a tool reading that contract, shows what you hold. Your wallet’s token list doesn’t. That distinction matters when deciding what to do with one. An inscription balance doesn’t automatically come with the wallet visibility or DEX compatibility people expect from an ERC-20. The marketplace works within that structure: listed inscriptions are locked, and payment and ownership change hands in one transaction. One detail I appreciate in the documentation is that sellers can cancel listings even while the platform is paused. There’s also a boundary worth understanding: the admin can upgrade the contract, so today’s rules aren’t a permanent guarantee. For me, understanding where ownership is recorded comes before choosing what to inscribe. It makes the rest of the product much easier to evaluate. The Paid Partnership label is used for transparency. This content is provided for informational purposes only and should not be considered financial or investment advice. Please conduct your own research and consider the risks involved.
🤝 Paid partnership
The Beepz Creator Program is here. Create content, tag @beepzfun, and climb the ranks through engagement. Top 50 creators share 10% of Beepz token creator fees + 10% Beepz inscription launch revenue. Plus marketing support for your own inscription launches on Beepz. Create. Climb. Earn. beepz.fun/creator
259
44
2
411
4,465
Marques Ken retweeted
Zerion’s biggest updates in October Perps trading, MegaETH support, etherfi in Earn, Kraken on Zerion API My breakdown 👇 1/ Product updates - Perps trading live for @zerion Premium (mostly iOS): 250+ assets incl. stocks, gold, silver, crypto, powered by @HyperliquidX - Fees cut: Standard 0.80% → 0.67%, Premium 0.40% → 0.25% - @megaeth fully supported: swap/bridge via @RelayProtocol, DeFi positions, Featured Apps - Portfolio app redesign: new sidebar, unified token selector, faster bridge - @ether_fi: weETH + Liquid ETH vault now inside Zerion Earn (Oct 12) 2/ The Kraken signal (Oct 20) @krakenfx published a case study: Kraken Wallet 2.0 runs on Zerion API Balances, DeFi positions, decoded history, PnL EVM + Solana My take: this is the strongest proof in the whole month A wallet built on your data layer is far stickier than a user who can delete the app in 10 seconds 3/ API updates for builders - Oct 22: counterparty_address added to each transaction act - Oct 22: /nft-positions endpoint, all NFTs by address, spam filter + floor price - Oct 19: full pagination for tx-subscriptions - Oct 8: application_metadata dropped when app is unknown - Oct 1: fewer duplicate share tokens (LP, aTokens, vaults) Small changelog items, but this is what keeps builders from leaving 4/ Stock tokens push - Oct 1 and 6: xStocks guides, trade 24/7 incl. earnings season - Oct 15: Robinhood Stock Tokens guide, 190+ stocks/ETFs Reminder: Robinhood’s are tokenized debt securities, not real shares, and US/Canada/UK/Switzerland are excluded 5/ A number that made me pause From the Sept Robinhood Chain report: 95% of Stock Token holders also hold speculative tokens So “tokenized equity investor” and “degen” are mostly the same wallet Great for Zerion’s volume Weaker for the “stable RWA flows” story 6/ Where I’d be careful - Most numbers (15M+ users, 42 chains) are self-reported - No independent coverage of Zerion in October - Zerion Alpha on Oct 9 was a September recap, so October’s core features aren’t summarized yet 7/ What I’m watching - Another big wallet publicly naming Zerion API after Kraken - Perps adoption beyond iOS Premium users - Early November Alpha recap 8/ My personal take What I like about Zerion is that it shows positions, not just balances LP, staking, lending all sit in one view instead of ten tabs October told me one thing They’re building the wallet as the storefront and the API as the business
51
8
46
7,245
Marques Ken retweeted
switching between AI agents, onchain cities and Kardashev energy concepts in a single day is definitely not the usual daily routine. ​that's honestly what's keeping the grind on @rarelistio interesting lately. ​having @fomopoly_rh , @agentdojo and @projectspherexy live at the same time gives you so much room to test different styles. ​not trying to overthink every post, just keeping the momentum going and building up RARS steadily.
57
15
86
1,224
Most agent setups live in one person's head. their prompts, their scripts, their way of fixing the failing build.. @superplanehq 1.0 turns that into a shared software factory for the whole team. same setup, same standards, and when someone improves the process everyone gets it ▸ picks tickets from your backlog that suit automation ▸ runs your agent and model of choice, retries through CI ▸ humans review the evidence before anything ships ▸ self-host it or use their cloud, Apache 2.0 52% of their own team's merged PRs ran through it last month
🤝 Paid partnership
Last month, 52% of the pull requests our team merged were automated through SuperPlane. Today, we’re launching @superplanehq 1.0: an open source software factory for engineering teams. We built it because using a coding agent can feel like becoming its assistant. Starting sessions. Checking its work. Chasing failing CI. SuperPlane takes on that coordination: - Scans your backlog and proactively picks up tickets suited to automation. - Runs a coding agent and verifies the result. Works through CI failures and review feedback. - Your team reviews the evidence and decides what ships. The setup, instructions, and standards are shared across the team. When someone improves the process, everyone benefits. Choose your agents and models. Run it on your own infrastructure. We spent more than a decade building CI/CD at Semaphore. We’ve seen how much teams can accomplish when routine work runs reliably in the background. Now we’re bringing that approach to coding agents. Here’s how it works:
107
35
113
3,141
Marques Ken retweeted
AI coding demos usually show you the fun part. prompt in, code out. but the real test starts after that. does it pass CI? does it survive review? how many times does someone need to step in before it's ready to merge? if people still have to keep pushing every step forward, faster code generation only solves part of the problem. maybe the bigger win is having agents handle that whole loop, not just write the code. 52% of @superplanehq's merged PRs were automated last month. so yeah, less time chasing PRs, more time building things.
🤝 Paid partnership
Last month, 52% of the pull requests our team merged were automated through SuperPlane. Today, we’re launching @superplanehq 1.0: an open source software factory for engineering teams. We built it because using a coding agent can feel like becoming its assistant. Starting sessions. Checking its work. Chasing failing CI. SuperPlane takes on that coordination: - Scans your backlog and proactively picks up tickets suited to automation. - Runs a coding agent and verifies the result. Works through CI failures and review feedback. - Your team reviews the evidence and decides what ships. The setup, instructions, and standards are shared across the team. When someone improves the process, everyone benefits. Choose your agents and models. Run it on your own infrastructure. We spent more than a decade building CI/CD at Semaphore. We’ve seen how much teams can accomplish when routine work runs reliably in the background. Now we’re bringing that approach to coding agents. Here’s how it works:
62
12
77
1,174
the part that stands out to me about @GenLayer is how it changes what a contract can actually judge imagine paying for a hotel booking through a smart contract, then finding out the room is non refundable even though the listing said otherwise a traditional smart contract can follow predefined rules perfectly, but it cannot independently read that listing and understand what the refund policy means, unless someone feeds it the relevant information an intelligent contract can fetch the booking page, use an llm to evaluate the terms, and determine whether the conditions are met, while validators independently check the result using their own models if they disagree, the protocol has ways to retry or allow an appeal that does not make web data automatically trustworthy, but it changes where we place our trust, from relying on one external answer to having a process for evaluating it to me, that is a much more interesting evolution than simply making smart contracts faster
Smart contracts need someone outside to tell them what happened. That's where decentralization ends. Intelligent Contracts go and check for themselves.
144
41
140
3,848
Marques Ken retweeted
If 88.5% is losing What are we still doing in this space lol 😂
total monthly users: 35m losing traders: 88.5% where’s all this money going?
141
50
1
205
53,543
Marques Ken retweeted
100 followers is all it takes to start?? that's a lower bar than i expected. who's actually going to try this
No cap on your commissions! Got 100 followers? It’s your time to shine with the Bitget Affiliate Program. • 600 USDT trading voucher pack • 40+% lifetime rebates • +10% boost in your first month

126
42
1
144
12,284
one thing i find interesting about @vangrid_io is that it doesn't need to own a fleet of robots or mapping devices to collect real world data a buyer can request a specific location, fund a bounty in usdc, and let someone nearby capture it with a phone the funds are escrowed onchain, the buyer reviews the submission, and an accepted capture moves into 3d reconstruction even ai agents can create these requests through x402, turning a software instruction into a real world data collection task that's a much more concrete use of crypto than adding a token to an app, the network coordinates demand, physical work, verification and payment in one flow the bigger question now is how far this model can scale as robotics and spatial ai need more data from places that change every day
272
44
1
256
15,319
Season 1 on @rarelistio is live! Join creator campaigns for curated projects and compete for a $50,000 prize pool. Verifying my account: RL-KR7LBH
2
3
74
Marques Ken retweeted
Day 3 of @Ptahtrade's Proof of Agent! 🤖 Today's estimated reward pool is $10,226.8 👀 Didn't get picked yesterday, so fingers crossed for Agent #13263 today! 🍀 I wish I had an AI agent that could catch opportunities while I'm sleeping. Crypto never sleeps, but I definitely need to 😂 What would you like AI agents to do for you in Web3? 🔥 #ProofOfAgent @Ptahtrade ptah.trade/poa
50
9
52
2,641
Marques Ken retweeted
privacy matters. and so does making onchain data useful without compromising it. primus is built exactly for that gap. zkTLS lets an attestor prove the authenticity of https or api data without revealing the raw content during the proof. fhe then allows computation on encrypted data so only the authorized party sees the result. institutions can verify what needs to be trusted (reserves, eligibility, activity) while keeping positions and strategies from being fully public. the data stays useful. the exposure stays limited. @primus_labs s.kaito.ai/Di33QnP points.primuslabs.xyz?ref=A0…
Privacy matters. And so does making onchain data useful without compromising privacy.
256
17
235
7,514