BTC $63,269, ETH $1,750.85, SOL $78.16, XRP $1.095, HYPE $67.05, DOGE $0.0731, AAVE $90.92.
The useful story this morning is value capture.
The last few digests already covered exchange-broker hybrids, tokenized stocks, regulated stablecoin issuers, wallet yield, and consumer-trust failures. Today’s cleaner signal is what happens after adoption arrives: who owns the ledger, the risk engine, the bridge, the vault, the power contract, and the regulatory category?
Swift launched a blockchain ledger for tokenized deposits with 17 global banks. JPMorgan warned that tokenization, payments, and settlement can grow in ways that barely reward public chains or their tokens. Chainlink CCIP has now pulled more than $7.2B of token value away from LayerZero after the exploit. Hyperliquid Policy Center and Phantom are telling the CFTC that onchain trading software should not be squeezed into old broker, exchange, and clearinghouse categories. Aave and Gauntlet are turning DeFi yield into institutional infrastructure. MARA is buying enough Texas power capacity to make its mining identity look secondary to AI data-center demand.
That is Friday’s read. Crypto adoption is no longer the hard part. The hard part is whether public networks, DeFi protocols, and token holders capture the economics after banks, exchanges, risk curators, and infrastructure vendors package the rails for normal finance.
Price snapshot via Coinbase spot prices and CoinGecko live market data around 04:10 HKT.
1. Swift’s Tokenized-Deposit Ledger Is The Clean Lead
The Block reported that Swift launched a blockchain-based shared ledger for a live pilot involving 17 banks across six continents.
The participant list is the story: Citi, HSBC, UBS, BNP Paribas, Standard Chartered, Wells Fargo, BNY, DBS, MUFG, and other global lenders.
This isn’t another stablecoin issuer fighting for exchange share. It is the bank messaging network giving banks a way to move tokenized deposits around the clock while preserving their existing compliance, credit, risk, and control systems.
That detail matters. Stablecoin advocates often assume faster settlement pushes value toward public blockchains. Swift is showing the opposite path: use blockchain design, keep the bank relationship, keep settlement tied to existing systems, and make the new ledger an orchestration layer instead of a public asset network.
If that model works, banks get 24/7 programmable money without handing the core payment relationship to crypto-native venues.
2. JPMorgan Put A Name On The Public-Chain Risk
The Block reported that JPMorgan analysts see a bigger structural risk than Strategy’s Bitcoin sale program: blockchain adoption that doesn’t benefit public chains or their tokens.
That is the right argument to lead with after the Swift news.
If tokenized deposits, tokenized funds, settlement workflows, and institutional payments mostly live on permissioned systems, crypto can win the adoption headline while losing the value-accrual fight. Activity rises, but fees, liquidity, governance, identity, and client relationships stay with banks and private infrastructure.
JPMorgan’s point isn’t that public chains disappear. The sharper claim is that they may become distribution and connectivity layers while regulated processing happens elsewhere.
That would hurt Ethereum more directly than Bitcoin, because Ethereum’s investment case leans harder on settlement usage, tokenized assets, stablecoins, and DeFi activity. Bitcoin can still trade as digital gold. Ethereum has to prove the public settlement layer remains worth paying for.
3. Chainlink CCIP Is Winning The Post-Exploit Bridge Rotation
CoinDesk reported that more than $7.2B of token value has migrated from LayerZero to Chainlink CCIP, with Mantle joining the move between July 9 and July 15.
This is one of the cleanest infrastructure rotations on the board.
The shift follows the $292M LayerZero exploit and now includes projects such as Kelp, Lombard, Solv Protocol, Virtuals, Re, Kraken’s tokenized assets, Yuzu Money, and Mantle’s MNT transfer system.
The useful signal isn’t “Chainlink good, LayerZero bad.” It is that cross-chain infrastructure has become board-level risk. Bridges no longer sit in the background as technical plumbing. They are capital rails, custody surfaces, governance dependencies, and brand risks.
When $7.2B moves because teams want a different security model, the market is voting on trust with migration work, not with tweets.
4. Hyperliquid And Phantom Are Drawing The Onchain-Regulation Line
The Block reported that Hyperliquid Policy Center and Phantom responded to the CFTC’s request for comment by arguing that onchain trading infrastructure needs its own rule treatment.
Their core argument is simple: legacy markets use brokers, exchanges, clearinghouses, and intermediaries that hold customer money. Onchain systems can let users keep control of funds while code handles execution and settlement.
That difference matters for Hyperliquid, Phantom, and every non-custodial trading interface. If writing software triggers the same obligations as running a traditional exchange or clearinghouse, the U.S. pushes serious onchain market structure offshore or into fully centralized wrappers.
The comment also lands while CME is fighting U.S.-regulated perpetual futures approvals and arguing about whether perps should be treated as swaps rather than futures.
The next CFTC fight isn’t only who can list perps. It is whether software, front ends, validators, market operators, and users get separate legal categories.
5. Aave Stable Vaults Turn DeFi Into A Back-End Product
The Block reported that Aave Labs released Stable Vaults for fintechs, wallets, exchanges, payment apps, and other institutional users that want to offer predictable stablecoin earnings.
This is DeFi becoming infrastructure rather than destination.
The vaults allocate deposits across Aave markets, Savings GHO, custom ERC-4626 vaults, chains, strategies, and approved bridges. The user doesn’t need to touch liquidity management, bridging, or strategy selection directly.
That is powerful distribution. It also changes the risk disclosure problem. A user may see a stablecoin savings product inside a familiar app while the back end moves through governance-approved strategies, bridge paths, claims, rebalancing, and multiple chains.
The product is less about DeFi users clicking around protocols and more about institutions embedding DeFi yield inside products people already use.
6. Gauntlet’s $125M Round Says Risk Curators Are Becoming Financial Infrastructure
The Block reported that Gauntlet raised a $125M Series C from SBI Holdings as the sole investor.
The investor pattern is the interesting part. SBI also put $76M into institutional crypto venue EDX earlier this week. That isn’t scattershot venture behavior. It is a push into the market-structure layer.
Gauntlet says it curates more than $1.5B in assets across vaults and works with more than 150 fintechs and institutions. It now plans to expand stablecoin coverage beyond USD and EUR into MXN, JPY, and other fiat-backed tokens.
This fits the same thesis as Aave Stable Vaults. DeFi yield is being repackaged into managed infrastructure, where risk models, optimization, curation, and fiat-stablecoin coverage become the product.
The protocol may get deposits. The curator may get the client relationship.
7. MARA’s 2 GW Texas Plan Makes Mining Look Like A Power Business
The Block reported that MARA shares rose roughly 14% after the company announced plans to acquire a 1,200-acre powered land site in Texas.
The site can support up to 2 gigawatts of grid capacity for high-performance computing and Bitcoin mining. The deal is structured as up to $600M in milestone-based payments tied to development progress.
This isn’t just a miner expansion story.
MARA is telling the market that power access is the asset. Bitcoin mining is one monetization path. AI data-center demand is another, and maybe the richer one if a tenant signs.
That changes how investors price mining companies. Hashrate matters, but power portfolios, interconnection rights, tenant pipelines, and development milestones may matter more.
The AI boom keeps turning miners into land, power, and data-center operators. Bitcoin mining is competing for its own infrastructure base.
8. Brazil’s B3 Added Options On BTC, ETH, And SOL Futures
CoinDesk reported that Brazil’s B3 exchange introduced options on bitcoin, ether, and solana futures on July 6.
This is the regulated-derivatives version of Latin America’s crypto adoption story.
The contracts settle into underlying futures rather than spot cryptoassets, so traders can hedge and express volatility views without custody, token transfer, or token administration.
That matters for local asset managers and traders who want crypto exposure inside a domestic regulated venue instead of offshore exchange accounts.
Brazil already has one of the more serious fintech and payment markets in the world. Options on crypto futures add another layer: not just access, but volatility management under a known market operator.
The next adoption step in Latin America may look less like retail stablecoin transfers and more like regulated derivatives desks.
9. Ethereum Institutional Wants Wall Street To Understand The Stack
CoinDesk covered Ethereum Institutional, a new nonprofit aimed at helping banks and financial institutions understand Ethereum.
The timing isn’t accidental.
JPMorgan is warning that permissioned infrastructure can bypass public chains. Swift is running tokenized deposits with banks. Tokenized funds and stablecoins keep expanding. Ethereum’s problem is no longer whether institutions are curious. It is whether they understand the public-chain case well enough to use it for more than distribution.
A neutral education layer can help, but it has to be honest. Wall Street doesn’t need another vague pitch about decentralization. It needs clear answers on privacy, finality, compliance, MEV, custody, validator risk, upgrade governance, transaction costs, and what public settlement can do better than a bank-owned ledger.
Ethereum has the deepest public settlement network. It still has to sell the part that banks can’t easily copy.
10. Bitcoin ETFs And Private Credit Are Flashing The Same Liquidity Warning
CoinDesk reported that redemption requests in the $2T private credit market reached $15.6B in the second quarter, while bitcoin ETFs also saw large outflows.
The pairing matters more than either number alone.
Crypto traders often read ETF flows as a Bitcoin-specific sentiment gauge. Private credit redemptions point to a broader risk-appetite problem: investors are asking for liquidity from products built around yield, duration, and sometimes less-transparent marks.
Bitcoin still bounced near $63K as markets looked past the latest Iran escalation. That resilience is useful. But it doesn’t erase the bigger signal.
When investors pull from both crypto ETFs and private credit, the market isn’t only debating Bitcoin. It is checking how much liquidity exists underneath crowded yield and risk trades.
11. GitHub Trending - Fresh Picks After The Repeat Filter
The repeat tracker ruled out the names featured from July 7 to July 9, including claude-video, karakeep, gastown, agent-skills, omnigent, T3MP3ST, Unlimited-OCR, MiMo-Code, ponytail, ai-job-search, OfficeCLI, and autoremesher. These three were clean enough to include today.
emilkowalski/skills has about 1.1K stars on Trendshift. It packages UI design guidance from the author’s writing into a skill file for designers and engineers. The signal is that agent quality is moving into reusable taste files: not just what command to run, but what good output should feel like.
vercel-labs/native has about 1K stars on Trendshift. It is a Zig-based toolkit for building desktop and mobile apps with web UI. The interesting part is Vercel pushing below the web app layer, where agents may need native shells, local permissions, and mobile packaging without giving up web-style iteration.
VoltAgent/awesome-design-md has about 989 stars on Trendshift. It collects DESIGN.md files based on major brand design systems. That is useful because coding agents keep improving at implementation, but still need explicit product taste, spacing rules, and component constraints to avoid generic output.
Morning Read
Read Swift’s tokenized-deposit ledger story, then read JPMorgan’s public-chain warning.
The number to remember is 17 banks.
That is enough to make the adoption story real, and enough to make the value-capture question uncomfortable.
Crypto has spent years proving that faster settlement, programmable assets, stable collateral, onchain risk systems, and global trading work. Now incumbents are packaging those ideas inside bank ledgers, regulated venues, managed vaults, risk-curation businesses, and power-backed data centers.
The public-chain side still has the better open network. It has to make sure the economics don’t leak away while everyone else copies the useful parts.
Evening Update - 20:05 HKT
BTC $64,235.54, ETH $1,788.96, SOL $79.14, XRP $1.11, HYPE $69.06, DOGE $0.0742, AAVE $97.68.
The evening update is deliberately not another pass over Swift’s tokenized-deposit ledger, JPMorgan’s public-chain warning, Chainlink’s CCIP rotation, the CFTC software-category fight, Aave vaults, Gauntlet’s round, MARA’s Texas power plan, Brazil’s B3 options launch, or Ethereum Institutional’s Wall Street brief.
The cleaner late-day signal is what breaks, changes, or becomes more serious after adoption leaves the pitch deck. Polymarket wants U.S. margin trading, which turns prediction markets from fully collateralized wagers into regulated leverage products. Metaplanet wants to turn its Bitcoin treasury into credit infrastructure. New Hampshire rejected the Bitcoin-backed muni bond right before launch. Robinhood Chain is already pulling ETH demand, but that demand is still tangled with memecoin behavior and Arbitrum beta. Ethereum Foundation is using AI agents for security work, and the hard part is no longer finding possible bugs. It is proving which reports are real.
That is Friday evening’s read: adoption is becoming a supervision problem. Once crypto products touch credit, margin, public finance, app-chain liquidity, protocol security, legal teams, issuer finance, manufacturing, and research benchmarks, the useful question changes. Who carries the risk when the product gets normal enough for real users?
Price snapshot via Coinbase spot prices and CoinGecko live market data around 20:05 HKT.
12. Polymarket Wants U.S. Margin Trading
CoinDesk reported that Polymarket’s U.S. affiliate applied for a National Futures Association license to offer margin trading to U.S. users.
This is a sharper comeback story than the trust campaign covered yesterday.
Fully collateralized prediction markets are easier to explain. A user posts the money, takes the position, and either wins or loses against a bounded payoff. Margin changes the product. It adds leverage, liquidation logic, user suitability, capital requirements, surveillance, and rulebook questions.
The timing matters. Kalshi received clearance for margin in March, prediction-market volumes reached $51B last year, and Bernstein expects the category to move toward $1T by 2030 if it keeps expanding across sports, crypto, politics, weather, and macro events.
If Polymarket wins this, the U.S. prediction-market fight moves beyond access. It becomes a market-structure fight over leverage on information itself.
13. Metaplanet Wants Bitcoin Treasuries To Become Credit Rails
CoinDesk reported that Metaplanet is studying Bitcoin-backed digital credit products with JPYC, Progmat, and Siiibo Securities.
This is the more interesting Bitcoin treasury story tonight.
Metaplanet holds 43,000 BTC, worth about $2.47B, and says the study could use Bitcoin as collateral for tokenized credit instruments with daily interest accrual, 24/7 trading, automated payments, and transparent redemptions.
That turns a treasury stack into a balance-sheet product.
The risk is obvious. Once Bitcoin collateral backs debt, holders care about valuation, liquidation, margin buffers, redemption mechanics, stablecoin rails, investor rights, and who eats losses if volatility arrives at the wrong time.
The opportunity is also real. Japan’s credit market favors large companies. If tokenized issuance can reduce operating burden for mid-sized firms without hiding risk, Bitcoin becomes more than a reserve asset. It becomes credit enhancement.
14. New Hampshire Killed The Bitcoin Bond At The Final Step
CoinDesk reported that New Hampshire’s Executive Council rejected the planned Bitcoin-backed bond project in a 3-2 vote.
That is a fresh consequence from yesterday’s municipal-credit story.
The proposed bond could have raised up to $100M for CleanSpark, used Bitcoin collateral, and already had a provisional Ba2 rating from Moody’s. The council vote was the final government approval step.
The rejection matters because it shows where Bitcoin collateral still scares public finance. A private company can carry BTC on a balance sheet. A state-linked bond has to defend reputation, accounting, liquidation design, investor protection, and political optics.
The bond may come back. Supporters are already arguing for reconsideration. But the failed vote is the real signal tonight: Bitcoin-backed public credit can make it through structuring and ratings work, then still die when elected officials decide the reputational risk is too high.
15. Robinhood Chain Is Already Creating ETH Demand
Cointelegraph reported that more than $70M of ETH has been bridged to Robinhood Chain in its first week, citing Token Terminal.
That changes yesterday’s Robinhood Chain read.
The first breakout story was CASHCAT, the memecoin that turned $800 into more than $1M. The fresh angle is infrastructure demand. Robinhood Chain is Arbitrum-based, EVM-compatible, and uses ETH as gas, so early usage can become a small but visible ETH demand source.
CoinDesk also reported that Arbitrum jumped as Robinhood’s onchain trading activity reached $568M.
That is the messy version of adoption. A regulated brokerage brand launches a chain for real-world assets. The early flows include ETH bridging, Arbitrum beta, tokenized-stock ambition, and memecoin speculation at the same time.
The chain may be serious. The users will still test it like crypto users.
16. Ethereum Foundation’s AI Security Work Makes Triage The Bottleneck
The Ethereum Foundation wrote that its Protocol Security team has been running coordinated AI agents against systems Ethereum depends on, including systems software, cryptographic code, and contracts.
The important detail is not that agents can generate bug reports. Everyone can generate bug reports now.
The Foundation said agents found real bugs, including a public CVE for a remotely triggerable panic in libp2p’s gossipsub, which Ethereum consensus clients rely on. It also stressed that most agent output is false positive noise unless humans and process can prove exploitability, severity, and fix scope.
This is the right security lesson for 2026.
AI does not make protocol security automatic. It raises the volume of plausible-looking findings. The scarce resource becomes senior triage, reproduction, disclosure discipline, and knowing which issue deserves emergency attention.
Ethereum’s agents found something real. The bigger product is the system that stopped the rest of the noise from wasting everyone.
17. Zcash Set Ironwood For July 28
Crypto.news reported that Zcash scheduled the Ironwood network upgrade for July 28 to replace the Orchard shielded pool after the counterfeit-token vulnerability.
This is a cleaner update from the privacy story earlier this week.
The old question was whether Zcash could prove its shielded pool had not hidden counterfeit supply. The new question is operational: can users, exchanges, wallets, miners, and infrastructure migrate cleanly before confidence decays?
Ironwood will require funds exiting Orchard to pass through an accounting checkpoint that could reveal whether counterfeit ZEC was ever created. That is a painful but necessary trade.
Privacy coins have a hard problem that transparent chains mostly avoid. Users want hidden amounts. Markets still need supply integrity. Zcash now has a date where that tension gets tested in production.
18. Coinbase’s Legal Chief Is Leaving After The SEC Fight
CoinDesk reported that Coinbase Chief Legal Officer Paul Grewal is stepping down and moving into an advisory role, with other legal leadership changes following.
The timing is the story.
Grewal was one of Coinbase’s public faces during the SEC fight. That legal battle shaped the company’s posture on exchange rules, token listings, staking, custody, market structure, and how aggressively a crypto company should fight regulators in public.
Now Coinbase is moving from litigation posture into product expansion: stocks, derivatives, prediction markets, trust-bank work, and international licenses.
That requires a different legal machine. The next phase is less about winning one enforcement narrative and more about getting many regulated products through many agencies without turning the app into a compliance maze.
Leadership turnover after a legal war does not mean Coinbase is weaker. It means the job changed.
19. Grayscale’s CFO Exit Shows ETF Issuers Are Still Under Pressure
CoinDesk reported that Grayscale CFO Edward McGee stepped down after seven years, with Kathryn Masci and Daniel Plourde named interim co-CFOs.
This belongs beside the Coinbase legal move.
Grayscale was once the default regulated Bitcoin access product. GBTC’s ETF conversion was historic, but it also exposed the firm to lower-fee competition. CoinDesk noted GBTC once held about $28.5B and now manages roughly $8.5B.
The company also paused IPO plans because of market conditions, according to a person previously cited by CoinDesk.
Crypto asset management is not just “institutions are here.” It is fee pressure, distribution fights, ETF flows, executive retention, public-market timing, and whether a brand built for one wrapper can keep its edge after the wrapper becomes crowded.
The ETF era did not remove competition. It made competition cleaner and harsher.
20. Bitdeer Is Bringing Mining Hardware Production To Nevada
Cointelegraph’s feed reported that Bitdeer shares rose after the company announced a $36M manufacturing facility in Nevada.
This is not another AI data-center lease story.
The plant is meant to produce SEALMINER Bitcoin mining hardware in the U.S., reduce reliance on third-party suppliers, and give Bitdeer more control over its infrastructure stack.
That matters because miners are splitting into two groups. Some are selling investors on power sites and AI hosting. Others are trying to own more of the mining supply chain itself: chip design, machine production, deployment, hosting, and treasury management.
Both paths are capital-intensive. Both claim strategic control. But they are different bets.
Bitdeer’s Nevada move says the hardware layer is still worth owning, even while the market is rewarding miners for becoming data-center landlords.
21. CoinDesk Research Says Q2 Was Crypto’s Third Straight Losing Quarter
CoinDesk Research published its Q2 2026 review, saying digital assets posted a third consecutive quarterly loss, the longest losing streak since the 2022 bear market.
That context matters because the day tape looks better. BTC pushed above $64K, ETH held near $1,790, HYPE moved back toward $69, and AAVE recovered toward $98.
But the quarterly read is colder. Institutional capital rotated into AI equities, Bitcoin ETFs saw their largest quarterly outflow since launch, and the market had to separate weak price action from continuing structural adoption.
That is the contradiction of 2026 crypto.
Tokenized assets, prediction markets, regulated brokerage products, stablecoin rails, public-chain security, and onchain credit keep moving forward. The liquid-token basket still has to prove it can capture that work in price.
Adoption can rise while returns lag. That is the uncomfortable theme traders should keep on the screen.
22. GitHub Trending - Fresh Picks After The Repeat Filter
The repeat tracker ruled out the names featured from July 7 through this morning, including claude-video, karakeep, gastown, agent-skills, omnigent, T3MP3ST, Unlimited-OCR, MiMo-Code, ponytail, ai-job-search, OfficeCLI, autoremesher, skills, native, and awesome-design-md. These three were clean enough to include tonight.
JustVugg/colibri has about 851 stars on Trendshift. It is a pure C local runtime for running a huge MoE model on a consumer machine by streaming experts from disk. The signal is local inference pressure moving below polished app layers into tiny runtimes that trade convenience for control.
malisper/pgrust has about 741 stars on Trendshift. It rewrites Postgres in Rust and says it now passes 100% of the Postgres regression tests. The useful angle is not “rewrite everything in Rust.” It is that AI-era infrastructure experiments are getting bolder around databases, safety, and long-lived systems.
stablyai/orca has about 15.3K stars on Trendshift. It is an agent development environment for running fleets of parallel coding agents across desktop and mobile. That fits the week: teams are moving from one assistant in one window toward orchestration, isolation, review, and coordination across several agents at once.
Evening Read
Read Polymarket’s margin application, then read Metaplanet’s Bitcoin-backed credit study.
The two stories rhyme.
Polymarket wants leverage on event outcomes. Metaplanet wants credit products backed by Bitcoin. New Hampshire rejected Bitcoin-backed public debt. Robinhood Chain is turning brokerage distribution into ETH and Arbitrum activity. Ethereum’s security team is learning that AI can find bugs but cannot replace triage. Zcash has to prove privacy can coexist with supply integrity. Coinbase and Grayscale are changing leadership as the legal and ETF battles enter a new phase.
The number to remember is $70M of ETH bridged to Robinhood Chain in one week.
That is enough to show real usage. It is also small enough to remind everyone that the next phase is still being built in public, with risk attached.