# bridge exploit — X 热门讨论 (2026-09-13 10:33 UTC)
## @0xchainink (Chain INK) · 09-13 10:23 · ♥25 ↻8 💬6 $CKB : Review 📜
What if you could give Bitcoin smart contracts without a bridge, without a wrapped token, and without asking Bitcoin to change a single line of code?
Meet Nervos CKB, a proof-of-work Layer 1 whose RGB++ protocol maps Bitcoin UTXOs directly onto its own Cell Model, so Bitcoin assets gain programmability through mathematics rather than through a custodian.
Designed by former Ethereum core developers, it can deploy post-quantum cryptography without a hard fork, and its Lightning layer is now in active testing.
Let's explore one of crypto's widest gaps between engineering and recognition. 👇
⚪ Nervos at a Glance
Marketplace Insight: Nervos occupies genuinely rare territory. It is one of the last proof-of-work Layer 1s of any scale, it uses a UTXO model rather than accounts, and RGB++ solves Bitcoin programmability through isomorphic binding rather than a bridge, which Messari singled out in a dedicated report. Crypto agility means post-quantum lock scripts can deploy as cells without a consensus vote, a structural advantage most chains cannot match. The clean contrast is that the recent 35% to 44% bounce is best read as a cheap, thinly traded Layer 1 rotating in a risk-on pocket rather than evidence Fiber usage has arrived. The engineering is respected and continuous. The demand has not shown up.
⚪ Mission
Nervos exists to give Bitcoin programmability without compromising it. The founding argument is that Bitcoin's security and its UTXO model are features rather than limitations, and that adding smart contracts should not require wrapping assets, trusting a bridge, or forking the base chain.
The Cell Model generalizes Bitcoin's UTXO into a container that holds arbitrary state, so RGB++ can bind Bitcoin UTXOs to CKB Cells cryptographically. State rent, where users pay for the storage they occupy, is the economic half of the design, intended to prevent the state bloat that eventually strains every other chain.
🔵 A Brief History
Nervos was founded in 2018 by a group drawn from the centre of the Ethereum ecosystem and China's early exchange industry, including Jan Xie, a former core R&D contributor to Ethereum, Daniel Lv, CTO of the imToken wallet, Terry Tai from the Yunbi exchange, Kevin Wang of Launch School, and Cipher Wang. That pedigree matters because the design choices were deliberate departures from Ethereum rather than ignorance of it: proof-of-work over proof-of-stake, UTXO over accounts, and state rent over unpriced storage.
The project raised $28 million from Sequoia Capital, Polychain, and Wanxiang, and mainnet launched in November 2019. For its first years, Nervos was a well-regarded Layer 1 that few people used, described by its own community as an outsider during the era of new smart contract platforms and rollups. The turn came with RGB++, launched in early 2024, which introduced a way to bind Bitcoin UTXOs to CKB Cells so Bitcoin-native assets like Ordinals and Runes could interact with smart contracts without a cross-chain bridge. It drove a 55% price increase, 181% month-over-month growth in new addresses that April, and a dedicated Messari report calling it a game-changer for Bitcoin's scalability.
The years since have been building against a falling chart. JoyID wallet passed 800,000 users and the RGB++ ecosystem grew past 400 dApps. The DAO 1.1 upgrade shipped, CKCON25 in December brought a new explorer and testnet releases, and Rosen Bridge integration passed a community vote in January 2026.
The setbacks were substantial. A $3.9 million bridge hack in June 2025 followed the earlier Force Bridge exploit, damaging trust and prompting a warning from South Korea's DAXA alliance, while KuCoin removed CKB from cross-margin trading in May 2026. The building continued regardless. Governance restructured around CKBA, a Swiss Verein replacing the old foundation coordination layer, with community votes passing Vellum, a reputation layer on did:ckb, at 64.37% and an iOS Pocket Node grant at 52.8%, while rejecting several proposals on merit.
The v0.209.0 node release hardened the tx-pool and networking stack, reaching roughly 89.5% miner adoption, and a new hard fork is in planning without a locked date. The busiest surface was Fiber: v1.1.0 shipped atomic Bitcoin Lightning swaps through a Cross-Chain Hub in August, Loop In/Out and hosted LSP work progressed, Android Pocket Node reached Google Play, and a July hackathon drew 66 submissions from roughly 100 participants, producing tools like Clasp for scoped, revocable AI-agent access to Fiber payments.
🔵 Ecosystem Narrative
The organizing idea is that Bitcoin should gain programmability through mathematics rather than through trust.
➛ RGB++ and isomorphic binding. Bitcoin UTXOs are mapped directly onto CKB Cells, so Bitcoin assets gain smart contract functionality with no bridge, no wrapped token, and no custodian, which is a structurally different security model from every wrapped-BTC approach.
➛ The Cell Model and CKB-VM. A generalization of Bitcoin's UTXO that can hold arbitrary state, paired with a RISC-V based virtual machine, giving the chain flexibility while keeping the UTXO security properties Bitcoin relies on.
➛ Fiber Network. A Lightning-compatible payment layer for RGB++ assets, with v1.1.0 enabling atomic swaps between Bitcoin Lightning and CKB through a Cross-Chain Hub, plus Loop In/Out, hosted LSP, and liquidity management in progress. Currently testnet-oriented with mainnet still pending.
➛ State rent economics. Users pay for the on-chain storage they occupy rather than paying once and storing forever, which prevents state explosion and gives CKB a demand driver tied to actual data usage.
➛ Crypto agility for post-quantum. Rather than requiring a hard fork to change cryptography, CKB allows new lock scripts including post-quantum ones to be deployed as cells without a consensus vote, so the chain can adapt to a quantum threat without forcing migration on anyone.
➛ Proof-of-work commitment. One of the last major Layer 1s still using Nakamoto consensus, which has retained a genuinely loyal mining community in a landscape that has largely moved to proof-of-stake.
⚪ Token Utilities
$CKB is a state-rent and security token rather than a conventional gas asset.
➛ State rent: holders pay for on-chain storage capacity, so occupying data space requires locking CKB, tying token demand directly to actual chain usage.
➛ Mining rewards: miners secure the proof-of-work network and receive primary issuance on a halving schedule capped at 33.6 billion.
➛ NervosDAO deposits: depositors lock CKB to receive a share of secondary issuance, which offsets the dilution that secondary issuance would otherwise impose.
➛ Transaction fees: pays for computation and transactions across the network and the RGB++ ecosystem.
⚪ Key Features
➛ RGB++ mapping Bitcoin UTXOs to CKB Cells with no bridge required.
➛ A Cell Model generalizing UTXO to hold arbitrary state, on a RISC-V virtual machine.
➛ State rent pricing storage to prevent state bloat.
➛ Crypto agility allowing post-quantum locks without a consensus vote.
➛ Proof-of-work consensus in a predominantly proof-of-stake landscape.
➛ Fiber Network extending RGB++ assets to Lightning-compatible payments.
🔵 Meet the Team
Nervos runs a notably lean current team, with three names listed formally against a protocol whose development happens largely through open-source contribution and CKBA-coordinated grants rather than a conventional company structure.
▶️ Core Members:
➛ Terry Tai - CEO and Director | A core developer of the Yunbi exchange, the venue that first listed Ethereum in China, before co-founding Nervos in 2018. He remains the formal executive lead through the transition to CKBA-coordinated governance.
➛ Kevin Wang - Co-Founder | Founded Launch School, an online software engineering education platform, after engineering at IBM's Silicon Valley Lab, and led the North American organization through the project's early years.
➛ Daniel Lv - Co-Founder | Former CTO of imToken, one of the earliest and largest Ethereum wallets with millions of users, bringing genuine consumer product experience to a project whose hardest problem has always been reaching users.
➛ The founding architects | Worth naming even though they no longer appear on the formal roster: Jan Xie, a former Ethereum core R&D contributor, designed the Cell Model and the deliberate departures from Ethereum that define the chain, while Cipher Wang architected RGB++ and drove the Lightning integration thesis. Much of what makes Nervos technically distinctive traces to their work.
➛ CKBA and CKB Eco Fund | The load-bearing entities today. CKBA, a Swiss Verein, replaced the older foundation coordination layer and now runs governance including grant votes and treasury design, while the Eco Fund finances ecosystem development. The project raised $28M from Sequoia Capital, Polychain, and Wanxiang.
🔵 Ratings
➛ Use Case: ★★★★ (4/5). Nervos solves a genuinely hard problem in a genuinely novel way. RGB++ binds Bitcoin UTXOs to CKB Cells cryptographically rather than through a bridge, which is a structurally safer approach than every wrapped-BTC alternative, and Messari published a dedicated report recognizing it. The supporting work is real: over 400 dApps, JoyID past 800,000 users, state rent solving state bloat at the economic layer, crypto agility allowing post-quantum locks without a fork, and Fiber extending RGB++ to Lightning with atomic swaps now working on testnet. The 1-point deduction is adoption relative to capability. The BTCFi narrative has not delivered the volume the architecture anticipated, Fiber remains testnet-oriented with mainnet pending, and competing Bitcoin layers have captured more attention despite weaker technical claims.
➛ Tokenomics: ★★★⯪ (3.5/5). The design is unusually thoughtful and solves problems most chains ignore. Primary issuance is capped at 33.6 billion on Bitcoin-style halvings with the next in November 2027, supply is roughly 98% circulating so there is no unlock overhang or vesting cliff, and state rent creates genuine demand tied to storage usage rather than speculation. NervosDAO lets depositors offset secondary issuance dilution, which is an elegant answer to perpetual inflation. The 1.5-point deduction is that secondary issuance of 1.344 billion annually is perpetual by design, network usage is too low for state rent or fees to offset it meaningfully, and the token remains roughly 97% below its peak despite the recent bounce.
➛ Audits: ★★★ (3/5). This is where the record is hardest. Nervos has suffered two bridge security incidents, including a $3.9 million hack in June 2025 that damaged trust and triggered a formal warning from South Korea's DAXA alliance, following the earlier Force Bridge exploit. Two failures on the same class of infrastructure is a pattern rather than an accident. The mitigants are genuine: the core proof-of-work chain has run since November 2019 without a protocol-level exploit, the UTXO Cell Model has a smaller attack surface than account-based systems, the v0.209.0 release shows continued security hardening across the tx-pool and networking stack, and crypto agility means cryptographic upgrades can ship without a contentious fork. But I could not verify a named third-party audit or security score, and the bridge record has to carry weight.
➛ Community: ★★★★ (4/5). The builder signal is stronger than the market cap suggests. A July Fiber hackathon drew 66 submissions from roughly 100 participants, producing genuinely creative work including Clasp for scoped, revocable AI-agent access to payments and a Cashu mint settling over Fiber, which is the clearest evidence of developer density this ecosystem has produced. Governance through CKBA functions on substance rather than rubber-stamping, passing Vellum at 64.37% and the iOS Pocket Node grant at 52.8% while rejecting proposals like fiber-payjoin-kit V2 and a mobile SDK on technical grounds. The miner base has held through a 97% drawdown, with 89.5% upgrading promptly to the current client, and community activity spans Kenya, Shanghai, Vietnam, and Colorado. The 1-point deduction is scale: roughly 70 CKBuilders and 45 tracked projects is a real ecosystem but a small one, and the DAXA warning and KuCoin delisting have narrowed mainstream access.
🔵 Conclusion
Nervos is one of the most intellectually serious projects in crypto and one of the least rewarded. A team drawn from Ethereum core development and the imToken wallet built a chain that deliberately reversed Ethereum's central choices, keeping proof-of-work and the UTXO model, then solved Bitcoin programmability through isomorphic binding rather than a bridge, which Messari recognized in print. State rent prices storage properly.
Crypto agility lets post-quantum cryptography deploy without a hard fork. Fiber is bringing Lightning-compatible payments to RGB++ assets, and a 66-submission hackathon shows builders are actually using it. The risks are real and partly self-inflicted.
Two bridge exploits, including a $3.9 million loss in June 2025, brought a regulatory warning from South Korea's DAXA and reinforced exactly the trust problem RGB++ was designed to solve. Secondary issuance continues perpetually while usage stays too thin to offset it. Exchange access has narrowed, Fiber remains testnet-oriented with mainnet pending, and the recent 40% bounce reflects a cheap asset rotating rather than demand arriving.
But the bull case has a specific shape. If Bitcoin programmability becomes genuinely important, the approach requiring no bridge and no custodian is the one that should win, and Nervos has been building it since 2018 with Lightning integration now in active testing and a developer community shipping real tools against it.
Nervos has spent eight years being technically right and commercially invisible. The question is whether the market ever arrives to check the work. https://x.com/0xchainink/status/2099081544375927152
## @changmyung1981 (오창명) · 09-13 03:18 · ♥21 ↻7 💬2 🧬 What if mitochondrial ATP does more than power T cells—what if it directly helps determine their identity by controlling chromatin accessibility?
A fascinating new Cell study provides a mechanistic bridge between immunometabolism and epigenetics, showing that mitochondrial ATP production can support the chromatin remodeling required for CD8⁺ T-cell effector differentiation.
The key molecule is D-α-hydroxybutyrate (DAHB).
Rather than functioning primarily as a fuel, DAHB acts as a metabolic signal that reprograms activated CD8⁺ T cells:
DAHB → OXPHOS ↑ → fatty-acid oxidation ↑ → phosphocreatine ↑ → local ATP buffering → BAF-dependent chromatin remodeling → effector genes ↑ → stronger antitumor immunity.
That is a remarkable metabolism-to-chromatin circuit.
🔥 DAHB pushes CD8⁺ T cells toward a cytotoxic phenotype
DAHB strongly increased:
Perforin IFN-γ TNF-family members GZMB CCL3/CCL4 and DAHB-pretreated OT-I cells showed approximately 3-fold greater tumor-cell cytotoxicity. But the metabolic mechanism was unexpected. DAHB increased ATP-coupled mitochondrial respiration while suppressing glycolysis. Estimated ATP production shifted substantially: OXPHOS-derived ATP ↑ ~82% glycolytic ATP ↓ ~23%.
Stable-isotope tracing showed that DAHB itself was not meaningfully oxidized through the TCA cycle.
Instead, it doubled fatty-acid oxidation and redirected glucose away from lactate production toward biosynthetic pathways, including the pentose-phosphate and serine/glycine pathways.
So:
DAHB is not primarily the fuel. It changes which fuels the T cell uses.
⚡ Then comes the central discovery: phosphocreatine
Untargeted metabolomics identified phosphocreatine (PCr) as the most strongly induced metabolite after DAHB treatment.
This is intriguing because the creatine–phosphocreatine system is usually associated with tissues such as skeletal muscle, heart and brain, where it rapidly transports and buffers high-energy phosphate.
Here, activated T cells appear to exploit the same system.
DAHB increased:
OXPHOS → mitochondrial ATP → PCr ↑ → CKB ↑
and blocking phosphocreatine production with cyclocreatine or creatine-kinase inhibition:
⬇️ perforin ⬇️ IFN-γ ⬇️ tumor killing.
Even more interestingly, CKB accumulated near the nuclear envelope, with DAHB producing >6-fold enrichment of CKB–Lamin B1 proximity signals.
This suggests that PCr may function as an intracellular energy shuttle:
mitochondria → PCr → nucleus → ATP
delivering energetic capacity precisely where ATP-consuming nuclear processes occur.
🧬 And one of those processes is chromatin remodeling ATAC-seq showed that DAHB increased chromatin accessibility at key effector loci: Prf1 Ifng Tnf
while inhibition of PCr synthesis largely reversed these changes.
The investigators then targeted the ATP-dependent BAF/mSWI-SNF chromatin remodeling complex.
Both BAF degradation and inhibition of its BRG1/BRM ATPase activity reduced DAHB-induced perforin and IFN-γ and eliminated the increased accessibility of effector loci.
This creates an elegant mechanistic sequence:
FAO ↓ mitochondrial OXPHOS ↓ ATP ↓ phosphocreatine ↓ nuclear ATP buffering ↓ BAF chromatin-remodeling activity ↓ open Prf1 / Ifng / Tnf loci ↓ CYTOTOXIC T-CELL DIFFERENTIATION
The graphical abstract on page 1 captures this beautifully: mitochondrial ATP is converted into a PCr-based energetic shuttle that connects mitochondrial metabolism to nuclear BAF activity and ultimately cytotoxicity.
🎯 The antitumor consequences were substantial Ex vivo DAHB conditioning of antigen-specific OT-I cells improved control of EL4-OVA tumors after adoptive transfer. DAHB also enhanced anti-CTLA-4 therapy in immunocompetent MC38 colorectal cancer and B16 melanoma models, whereas this benefit disappeared in lymphocyte-deficient RAG2-KO mice—supporting an adaptive immune mechanism.
Figure 6 on page 11 is especially compelling: DAHB improves tumor control across adoptive-transfer and checkpoint-blockade experiments.
And the TIL phenotype is interesting.
DAHB-treated tumors contained CD8⁺ T cells with:
⬆️ CD44 ⬆️ TCF1 ⬆️ antigen-specific gp100⁺ cells ⬇️ PD-1⁺CD39⁺ cells ⬆️ polyfunctional IFN-γ⁺TNF-α⁺ responses ⬆️ T-bet.
Importantly, residual PD-1⁺CD39⁺ cells expressed more TCF1, suggesting enrichment toward a precursor-exhausted rather than terminally exhausted state. 👨⚕️ And the biology translated to human T cells Under chronic stimulation, DAHB-treated human CD8⁺ T cells maintained: IFN-γ ↑ TNF-α ↑ PCr ↑ ATP/ADP better preserved and cyclocreatine partially reversed these effects. Human CD8⁺ T cells engineered with the NY-ESO-1-specific 1G4 TCR also showed greater cytokine production and stronger killing of A375 melanoma cells after DAHB treatment. Most importantly, in an A375 xenograft model: DAHB alone → little effect on tumor growth but 1G4 T-cell ACT + DAHB → significantly improved tumor control. That distinction argues that DAHB is functioning primarily as a metabolic adjuvant for T cells, rather than simply acting directly on cancer cells. 💡 The conceptual implication is broader than DAHB We often describe mitochondrial metabolism as providing enough ATP for cells to “function.” This study suggests something more specific: Energy availability can regulate cell fate because chromatin remodeling itself has an energetic cost.
ATP is therefore not merely downstream fuel for an already differentiated effector T cell.
Mitochondrial energy production may participate directly in establishing the epigenetic accessibility required to become and remain an effector cell.
That has potentially important implications for:
CAR-T manufacturing TCR-engineered T cells tumor-infiltrating lymphocytes checkpoint blockade metabolic preconditioning of adoptive cell therapies.
Instead of engineering only receptors and signaling domains, perhaps we should also engineer the bioenergetic infrastructure that maintains their chromatin state.
There are important caveats.
The molecular sensor through which DAHB activates mitochondrial ETC activity remains unknown. The concentrations used experimentally were high, whether microbiota-derived DAHB ever reaches comparable concentrations in vivo is unclear, and the safety/tolerability of therapeutic systemic DAHB in humans remains untested.
Still, the mechanistic concept is striking: MITOCHONDRIA → ATP → PHOSPHOCREATINE → BAF → CHROMATIN ACCESSIBILITY → T-CELL FATE → ANTITUMOR IMMUNITY Metabolism and epigenetics are not separate layers of T-cell biology. They may be connected by something as fundamental as where ATP is produced—and where that energy is delivered. 📄 Ng C, Fung TS, Li D, et al. Mitochondrial ATP promotes T cell differentiation through chromatin accessibility. Cell. 2026;189. DOI: 10.1016/j.cell.2026.08.023
#CancerImmunology #Immunometabolism #TCells #Mitochondria #OXPHOS #Phosphocreatine #Chromatin #Epigenetics #BAF #TCellExhaustion #Immunotherapy #CART #CellTherapy #CancerMetabolism https://x.com/changmyung1981/status/2098974646020530338