橋樑之戰：鏈抽象協議會成為 Web3 互通性的未來嗎？

現時用戶經常要同時處理大量孤立的區塊鏈——每個都有其專屬錢包、代幣及應用程式——令跨鏈操作變得繁瑣。這種分割被廣泛視為大眾應用的一大障礙。事實上，有分析人士也指出，「用戶與流動資金分散在不同區塊鏈，日益構成 Web3 應用的阻力」。

當每條鏈都像獨立孤島時，移動代幣或數據一般需繁複的橋樑或手動兌換，成本高又有安全風險。例如，2022 年針對跨鏈橋樑的攻擊就佔了當年被盜加密貨幣總額約 69%，突顯傳統橋樑的高風險。

鏈抽象協議旨在為用戶帶來順暢體驗，讓他們無需管理多個錢包或理會代幣在哪條鏈上。實際上，用戶可像全部錢包餘額和資產已合而為一那樣操作。鏈抽象的本質是「移除 UX 阻力」，讓用戶跨鏈操作時，「就像只有一條鏈一樣」使用錢包。即是說，協議會自動在後台處理用戶需手動選鏈及支付不同費用的煩瑣。

以下會探討鏈抽象協議的原理、與傳統橋樑的比較，以及這種新方法是否真有機會成為下一代 Web3 互通方案。

傳統橋樑的局限

對很多早期用戶來說，橋樑是解決區塊鏈分割的第一步。跨鏈橋本質上是一種連接兩條或以上區塊鏈的專門應用，允許代幣或數據從一條鏈轉至另一條。傳統橋樑一般做法是：在來源鏈鎖定（或銷毀）代幣，在目標鏈鑄造（或解鎖）等值代幣，從而「橋接」價值。例如你想把 ETH 由以太坊轉至幣安智能鏈（BSC），可以把 ETH 存入以太坊的橋樑合約，然後橋樑會於 BSC 鑄造對應 ETH。這種方式確實帶來更多功能——用戶可搬運資產到不同鏈，觸及不同流動性。

不過，橋樑操作需用戶主動多步操作，還有不少取捨。用戶通常要選擇合適的橋樑，指定來源和目標鏈，每條鏈獨自管理 gas 費，有時還要等待多次鏈上確認。Arcana Network 形容這作為「推式」橋樑，用戶要負擔大部份流程：逐步選橋、選來源目的地、逐項進行操作，每一步都增加複雜性和延誤。例如，把資產傳到鏈 A 的橋後，往往要在鏈 B 獲得簽名，然後手動於目標鏈認領或解鎖代幣。任何鏈的網絡壅塞或確認延誤都會加長過程。

安全亦是橋樑一大隱患。橋樑通常依賴特殊合約或一組驗證者保管資產或驗證操作，容易成為駭客攻擊目標。Chainalysis 指出 2022 年約 70% 加密貨幣盜竊來自橋樑攻擊。著名如 Wormhole 被駭（2022 年 2 月損失 3.21 億美元）等橋樑意外，突顯中心化漏洞點會被戲劇性利用。只要橋樑有一個驗證節點或金鑰被盜取，或者智能合約有漏洞，攻擊者就可盜走資金。Chainalysis 警告「跨鏈橋……通常存在資金集中存儲點，成為目標」。

用戶和開發者亦會感受「橋樑之痛」。每增一條鏈，就多一個（甚至多個競爭）橋樑，錢包與代幣數量激增。跨鏈搬資產或執行合約，用戶需希望被要求完成繁瑣多步動作。這種分割令資金和用戶都被孤島化：每條鏈基本上留住自己的社群，除非願意承受麻煩搬鏈。因此不少用家乾脆只用單一鏈，盡量避免橋樑——而這種分隔正是互通性要消除的。總結而言，橋樑可行，但代價是：複雜、延遲和高安全風險。

鏈抽象協議登場

鏈抽象協議則採取根本不同做法。與其逐筆逐步由用戶主動發起跨鏈操作，這些系統主張「拉式」模型：用戶只需指定意圖或最終目標，協議基礎設施會自動處理所有跨鏈細節。在高層次上，鏈抽象即是「徹底簡化用戶體驗與加強互通性」，把多鏈操作的複雜包裝起來。實踐上，通常意味用戶可用同一介面或帳號，連接任何支援的區塊鏈，而背後的智能機制會自動安排跨鏈交易與資金流。

舉個例子：用鏈抽象跨鏈發送代幣，用戶無需自己操作橋樑，只需指令跨鏈應用如「將 100 USDC 發給 Bob 在 Z 鏈」。鏈抽象層（常見靠智能合約與鏈下轉發者 relayer）會判斷如何實現這個意圖：找最佳路徑、必要時兌換代幣、在適當鏈支付 gas，最後送到對方手中——過程全自動，用戶無須理會細節。用戶表面只看到單一「發送」操作，實際上抽象層可能同時執行多次橋樑操作。這種「基於意圖」流程是許多鏈抽象設計的核心特色，讓開發者和用戶當多鏈如單鏈操作。

鏈抽象直攻 Web3 兩大痛點。第一，大幅簡化用戶體驗（UX）。用戶無需分開管理不同鏈的錢包或 gas 代幣。只需一個統一帳號或介面，像所有區塊鏈合一運作。Blockworks 說鏈抽象能「把區塊鏈基礎架構從用戶體驗中脫鉤」，多鏈互動「成為背景細節，用戶基本無感」。在這模式下，用戶登入 dApp 一次，應用便能偷偷在背後完成以太坊、Polygon 或 Avalanche 等交易——用戶不用手動切換網絡或用多個錢包。這個跨鏈體驗常被比擬為今天的互聯網：你可瀏覽及使用服務，根本不需要知道底層協議。

第二，鏈抽象打破流動性與開發者分割。在零碎系統下，開發者往往要選擇單一條鏈（用戶及資產有限），或為每條鏈分開開發版本。鏈抽象則讓 dApp 能同時觸達所有支援鏈上的用戶及流動資金。Blockworks 說，這做法令流動性「成為全球商品」，不再被個別鏈鎖死。假如 DeFi 協議用鏈抽象，可以自動於多鏈池找最佳利率。

開發者重心可轉為功能創新，而非為不同鏈重複移植程式。這種「網絡效應」可帶動加速擴展；Blockworks 更認為，讓應用可「直接存取整個生態內的價值和用戶」，有助最佳化成本與可擴展性，例如可按應用不同部分，挑選最合適鏈：A 鏈借貸便宜、B 鏈兌換流動性深，協議會自動分配——全部在單一用戶體驗下完成。

簡單說，鏈抽象將多鏈混亂化繁為簡，建構單一邏輯鏈。用戶用一個「統一」錢包即可一步完成跨鏈兌換或消息，而開發者部署 dApp 毋須針對每鏈重寫程式。如果理念實現，協議勢將帶來傳統區塊鏈初衷的易用性：消費者如用普通網絡應用一樣與區塊鏈互動，不必知道背後牽涉哪些鏈。

鏈抽象協議運作原理

雖然細節有異，大部份鏈抽象協議架構相近。大致有三層（可稱許可層、求解層、結算層），共同處理用戶「意圖」由頭到尾。

許可層（帳戶抽象）：常以智能合約錢包或帳戶抽象機制實現，作為通用帳戶。用戶用同一帳戶簽署並指定意圖（如「把代幣轉去 Z 鏈」）。帳戶抽象層可驗證請求並轉交給求解網絡，也可統一處理 gas（讓用戶可用任何代幣付 gas，甚至預付）。
求解層（鏈下轉發者或「解決者」）：獨立節點或服務網絡，競爭處理用戶意圖。每有意圖送交，多個解決者可競標處理（有時需質押保證金）。解決者會組合多鏈交易以實現意圖，若需可先墊付資金。它們本質像專業經紀或做市商。例如，Alice 想將 X 鏈的 A 代幣兌換成 Y 鏈的 B 代幣，解決者會借入 X 鏈 A 代幣，... on Y, and deliver it – all atomically, then claim payment for its service. Users may pay a tiny fee or the solver takes a spread. This off-chain solver model is what enables the “pull” approach: users don’t need to find or trust any single bridge themselves – they just define their goal.

在 Y 上，一次性完成並交付整個操作，然後申領服務費。用戶可以支付小額手續費，或者由解決者從中賺取差價。這種鏈下解決者模式正是實現「拉式」方法的關鍵：用戶無需自己尋找或信任單一橋接器——他們只需定義目標。

Settlement layer (Cross-chain execution): Once a solver has packaged the required transactions, they use actual cross-chain messaging or bridging infrastructure to execute them. This might involve locking tokens in a bridge contract, using an oracle protocol like CCIP, or even using built-in light-client validation. The settlement ensures the end result: tokens have moved and final blockchain states match the user’s intent. Security at this layer still depends on the underlying mechanism (multi-sigs, proof networks, etc.), but because the user offloads complexity, it often looks like just one seamless transaction from their perspective.
結算層（跨鏈執行）：當解決者包裝好所需交易後，會利用實際的跨鏈訊息傳遞或橋接基礎設施去執行交易。這可能包括將代幣鎖定於橋接合約，使用如 CCIP 的預言機協議，甚至利用內置的輕客戶端驗證。結算層確保最終結果——代幣成功轉移，區塊鏈的最終狀態吻合用戶原先意圖。這一層的安全性依然取決於背後的機制（多簽、證明網絡等），但用戶因為交由協議處理複雜步驟，體驗上會像進行了一次簡單無縫的交易。

Chain abstraction protocols differ in how each layer is implemented. Some, like LayerZero and Axelar, focus on lightweight messaging protocols for the settlement layer. Others, like Across or Router, emphasize the solver network and economic incentives. Chainlink’s CCIP is an oracle-based messaging standard, while Hyperlane (formerly Abacus) is focused on permissionless, modular cross-chain message passing. Some newer designs (sometimes called “Intent-based” protocols) may incorporate advanced features like flash executions or integrated swaps.

各種鏈抽象協議在每一層的實現方式上各有不同。有些，例如 LayerZero 和 Axelar，主力於結算層的輕量級訊息傳遞協議。亦有如 Across 及 Router，重點在於解決者網絡及經濟誘因。而 Chainlink 的 CCIP 屬於預言機訊息標準，至於 Hyperlane（前稱 Abacus）則針對無需許可的模組化跨鏈訊息傳播。一些新型設計（有時稱為「意圖驅動」協議）甚至引入進階功能，例如閃電執行或內置兌換。

What they share is the goal of shifting work from users to the protocol. Instead of a user manually bridging assets, these systems orchestrate the operations across chains automatically. For instance, if a user simply indicates “send 5 USDC to Chain B,” the chain abstraction service might handle swapping any USDC subtype, paying gas on Chain B, and then delivering the tokens, all invisibly. The user never sees intermediate steps.

它們的共同目標是將繁重的操作由用戶轉移到協議層。用戶毋須親手跨鏈轉賬，系統會自動安排跨鏈操作。例如，用戶只需表示「發送 5 USDC 去 Chain B」，鏈抽象服務可以悄悄在背後完成兌換不同類型 USDC、在 Chain B 支付 gas 費，並把代幣送到指定地方。中間過程對用戶全都隱藏，完全無縫。

Bridges vs. Abstraction: A Comparison

橋接 vs. 抽象化：比較

To understand the “battle” between bridges and abstraction, it helps to compare their trade-offs directly. Arcana Network’s analysis frames this well. Traditional bridges use a push model: users actively push their assets into a bridge, specify source and destination, and then manually claim on the other side. Chain abstraction uses a pull model: the user issues an intent (a destination and an action) and the system pulls funds and executes on their behalf.

要理解橋接與抽象化兩者的「角力」，直接對比其取捨會很有幫助。Arcana Network 的分析對此有很好的說明。傳統橋接屬於推式模式：用戶主動將資產推入橋接，使出發地及目的地，之後需在對方鏈上手動認領。而鏈抽象則用拉式模式：用戶發出意圖（目標和行動），系統自動拉取資金並代表用戶執行。

In practice, this means bridging can give the user full control and choice (which bridge contract to use, exact chains, etc.), but at the cost of complexity and multi-step procedures. Chain abstraction, by contrast, simplifies steps: the user only chooses the final outcome, and the protocol “pulls” the right chains and fees. Arcana notes, for example, that with chain abstraction “the abstraction layer handles selecting the appropriate chains, managing gas fees, and executing the transaction through intents,” whereas with a regular bridge the user must manually handle each part.

實際上，橋接給予用戶更大自由——可以自行選擇橋合約、確定來源/目的鏈等——但卻更加繁複，需要經歷多個步驟。相比之下，鏈抽象讓程序簡化：用戶只選定最終結果，其餘如選鏈、處理手續費等都由協議「拉取」處理。Arcana 指出，鏈抽象情境下，抽象層負責幫用户選合適的鏈、管理 gas 費、根據意圖執行交易；反觀普通橋接，一切都要用戶逐步親自操作。

Security considerations also diverge. Bridges, by design, often centralize risk (a locked vault, multisig, or set of validators), making hacks lucrative. Yet they can be made relatively simple and specific for asset transfers. Chain abstraction protocols reduce direct user exposure by handling chains through relayers. Arcana argues that this “reduces the attack surface area by minimizing direct user interactions with multiple chains”. With abstraction, the user’s wallet doesn’t directly call every chain’s contract; instead, it authorizes a single abstract transaction. However, this puts more trust in the abstraction layer’s infrastructure (the relayers and oracles). If those are compromised, cross-chain actions can still fail or be hijacked. In short, both models require careful design: bridging has historically been costly when it fails, but abstraction still depends on robust protocol security.

兩者於安全設計上亦有明顯分別。橋接通常風險集中化（鎖倉、多簽、驗證者等），容易成為駭客目標，但轉資產機制相對簡單明瞭。鏈抽象協議通過中繼者代為處理跨鏈，減低用戶直接暴露。Arcana 認為這「減少攻擊面，因爲用戶毋須直接與多條鏈互動」。用戶的錢包不需要直接與每條鏈的合約互動，只需授權一次抽象交易。不過這樣便更依賴抽象基建（如中繼者及預言機）。這些基礎設施一旦被攻擊，跨鏈行動仍有可能出問題甚至被劫持。總結而言，兩種模式皆需審慎設計：傳統橋接失敗時歷史上往往損失巨大，而抽象模式依然非常依賴協議自身的安全性。

Performance and cost also shift. Bridges require two on-chain transactions (lock and unlock) plus confirmations on each chain, so they can be slow (often minutes or more). Chain abstraction can be faster by using atomic or batched operations: a single user request can trigger fewer user-visible steps. Solver networks can optimize routes to avoid congested chains and even negotiate fees. Arcana points out that chain abstraction’s competitive solvers can dynamically find the fastest, lowest-cost paths, potentially making swaps or transfers more efficient. In some designs, multiple solvers bid to complete an intent, pushing costs down.

兩者在性能和成本上差異亦大。橋接要進行兩次鏈上操作（鎖倉、解鎖），並需各鏈多次確認，通常需時數分鐘或更長。鏈抽象則可透過原子性或批量操作加快流程：用戶只發出一個請求，系統背後可大幅減少顯性步驟。解決者網絡還能優化路徑，避開壅塞鏈，甚至協商更優惠的費用。Arcana 指出，鏈抽象中的解決者會競爭性地尋找最快、最便宜路徑，令兌換/轉賬效率大增。有些設計更允許多個解決者競價執行意圖，進一步壓低成本。

However, chain abstraction systems are more complex architecturally. They often involve off-chain components, staking incentives for solvers, and sometimes novel economic schemes. This complexity can make them harder to analyze and build trust in. Bridges, in comparison, are relatively simple smart contracts (though as we’ve seen, that simplicity can invite hacks).

然而，鏈抽象系統架構更為複雜，通常涉及鏈下組件、解決者質押等經濟誘因，甚至新穎的 tokenomics。這種複雜度令其分析和建立信任更難。相比之下，橋接一般是相對簡單的智能合約（但正如我們見到，越簡單有時越容易被駭）。

In summary, traditional bridges offer direct but cumbersome cross-chain transfers, while chain abstraction offers a sleek UX at the cost of extra protocol layers. The key question is whether users and developers will trade simplicity for new infrastructure trust. As one researcher put it, chain abstraction is meant to simplify UX and interoperability, “but the underlying process and technology” remain intricate.

總結而言，傳統橋接提供了直接但繁瑣的跨鏈轉帳方式；鏈抽象則以更多協議層為代價，換來流暢簡潔的用戶體驗。關鍵課題在於：用戶及開發者會否願意以信任新基建來換取使用簡單？正如有研究者所說，鏈抽象是為了簡化 UX 及促進互通，「但其底層過程和技術依然極為複雜」。

Notable Chain Abstraction Projects

重要鏈抽象項目

A variety of projects are pioneering chain abstraction and related interoperability. Many started as bridges or messaging protocols and evolved.

市面上有眾多項目正在推動鏈抽象及跨鏈互通，許多本是橋接或訊息協議演化而來。

Key examples include:

重點例子包括：

LayerZero: One of the most-discussed abstraction layers, LayerZero provides an “omnichain” messaging protocol. It uses an oracle-and-relayer design: each endpoint chain runs a light node and trustlessly checks messages from the other chain, while an off-chain relayer (like Chainlink) helps transmit the data. This design lets contracts communicate directly across chains without needing a two-step transfer. LayerZero co-founder Sandeep Nailwal describes it as a “lightweight cross-chain information transmission” system where security is guaranteed by both chains verifying each other’s messages. It has been integrated into many protocols, allowing them to support multiple chains seamlessly. LayerZero aims to abstract the choice of network away from users, providing a unified messaging layer for DApps.
LayerZero：作為當前最熱門的抽象層之一，LayerZero 提供「全鏈」訊息協議。它採用預言機 + 中繼設計：每個端點鏈都運行輕節點並驗證對方鏈發來的訊息，而鏈下中繼者（如 Chainlink）則協助傳遞數據。這種設計允許合約可直接跨鏈溝通，不需要兩步轉帳流程。共同創辦人 Sandeep Nailwal 形容其為「輕量級跨鏈資訊傳遞」方案，保障由兩條鏈互相驗證訊息來實現。目前已被眾多協議納入，輕鬆覆蓋多鏈。LayerZero 的目標是在用戶體驗上完全抽象出網絡選擇，為 DApp 提供統一訊息層。
Axelar: Built with Cosmos technology, Axelar is a decentralized network of validators that connects many blockchains. Axelar advertises “full-chain deployment,” meaning a DApp built on Axelar can run on any supported chain out of the box. Axelar’s interchain protocol lets developers deploy one application and automatically have it available on dozens of networks. Under the hood, Axelar validators participate in cross-chain API calls and token transfers, enabling abstracted commands like “transfer USDC from BSC to Ethereum” with one function call. Because of its broad coverage, Axelar claims to support the most public chains of any abstraction layer. This wide reach helps both user experience (fewer choices to make) and liquidity (aggregating tokens across chains).
Axelar：基於 Cosmos 技術構建，Axelar 是個去中心化驗證者網絡，可連接多條區塊鏈。Axelar 標榜「全鏈部署」，意即用戶只需在 Axelar 上構建 DApp，便可即時應用於所有支援的鏈上。其跨鏈協議使開發者部署一個應⽤即可自動支援數十條網絡。實作上，Axelar 驗證者參與處理跨鏈 API 調用及代幣轉移，使開發可用如「一鍵把 USDC 由 BSC 轉至 Ethereum」的高層抽象命令。涵蓋範圍之廣，使 Axelar 號稱支持最多公共鏈的抽象層，有利於用戶體驗（選擇更少）、亦有助於流動性（集中多鏈資產）。
Wormhole: Originating as the bridge between Solana and Ethereum, Wormhole has expanded into a network of “core contracts” and guardian nodes that connect multiple chains (Solana, Ethereum, Terra, BSC, etc.). A trusted quorum of guardians signs off on messages; each blockchain runs a Wormhole contract that emits or reads these signed messages. Wormhole functions both as a token bridge and a generalized messaging fabric. Because Solana was one of the initial supported chains, Wormhole has quickly accumulated high volume. However, it too relies on guardian signatures (and was the target of one of crypto’s largest exploits). Wormhole represents an intermediate model: more decentralized than a single bridge contract, but still a multi-sig system.
Wormhole：最初作為 Solana 與 Ethereum 之間的橋接器，Wormhole 逐漸演化為一個由「核心合約」及守護節點組成、連接多鏈（Solana、Ethereum、Terra、BSC 等）的網絡。一組守護節點組成的可信多數負責簽名訊息，每條鏈都運行一份 Wormhole 合約，生成或讀取這些被簽名的訊息。Wormhole 既是代幣橋，也是通用訊息層。由於最初支援 Solana，Wormhole 很快吸納大量量能。但同樣依賴守護者簽名（也曾成為史上最大型加密攻擊目標之一）。它屬於中間型設計：比單一橋更去中心化，但本質上仍為多簽系統。
ZetaChain: Launched more recently, ZetaChain is a Cosmos-SDK chain that natively runs smart contracts which can operate on multiple networks. It’s essentially a standalone blockchain designed for interoperability. ZetaChain’s goal is “full-chain smart contracts,” meaning a dApp running on ZetaChain can interact simultaneously with Ethereum, Bitcoin, and other chains within one contract’s logic. As the Binance summary notes, ZetaChain is similar to Axelar in architecture but explicitly includes Bitcoin (via a special node) and targets on-chain composability across all chains. It’s an ambitious example of building abstraction into a new layer-1 blockchain.
ZetaChain：作為較新推出項目，ZetaChain 是一條 Cosmos-SDK 鏈，可原生運行跨多條網絡的智能合約。這是一條以互通為目標、自成一格的區塊鏈。其目標是「全鏈智能合約」，即一個運行於 ZetaChain 的 dApp 可以同時通過單一合約邏輯連接 Ethereum、Bitcoin 等多條鏈。如 Binance 簡介所說，ZetaChain 架構上與 Axelar 相近，但特別支援 Bitcoin（通過特殊節點），重點實現鏈上合成式組合。這是在新 Layer-1 打造抽象層的進取例子。
Chainlink CCIP (Cross-Chain Interoperability Protocol): Rather than its own network, Chainlink is extending its oracle framework into interoperability. CCIP is an open standard that uses Chainlink’s decentralized oracle and its new Risk Management Network to relay messages and tokens between chains. In essence, it provides a universal bridge mechanism accessible to any developer. According to Chainlink, CCIP will underpin tools like a “Programmable Token Bridge” to move tokens across any blockchain in a “highly secure, scalable” way. Because it leverages Chainlink’s large operator pool, CCIP promises robust verification and a built-in dispute/prevention layer. It aims to hide cross-chain intricacies by letting developers make one CCIP call instead of multiple bridged steps.
Chainlink CCIP（跨鏈互通協議）：Chainlink 並非獨立運營網絡，而是把本身預言機技術擴展至互通場景。CCIP 是一個開放標準，利用 Chainlink 去中心化預言機及新搭建的風險管理網絡，在鏈間傳遞消息與代幣。簡單來說，它提供一個可供所有開發者調用的通用跨鏈橋。根據官方，CCIP 將成為如「可編程代幣橋」等工具的底層方案，可「高度安全、可擴展」地在任何區塊鏈間轉移代幣。因採用 Chainlink 的大型運營節點池，CCIP 可確保強大驗證及內置糾紛/預防層。其目標是讓開發者只需調用一次 CCIP，而不必多次調用具體橋接步驟，把跨鏈細節全隱藏。
Hyperlane (formerly Abacus): Hyperlane provides a modular cross-chain messaging system that can be self-hosted by any chain or app. Unlike LayerZero’s chain-native design or Axelar’s validator network, Hyperlane lets developers self-deploy a Hyperlane instance on their chain to connect to others. It emphasizes “permissionless expansion”: once one uses Hyperlane on a given blockchain, it automatically gains connectivity to all chains already using it. Hyperlane reports that it has “140+ chains connected” and has bridged over $8 billion across its network. It supports multiple virtual machines and encourages users to “own your interop”
Hyperlane（前稱 Abacus）：Hyperlane 提供可模組化的跨鏈訊息系統，任何鏈或應用都可自我部署。與 LayerZero 的鏈原生設計或 Axelar 的驗證者網絡不同，Hyperlane 允許開發者在自己的鏈上隨時搭建 Hyperlane 節點，即可自動連接所有其他 Hyperlane 節點。強調「無需許可擴展」：只要某鏈啟用 Hyperlane，即自動與所有已用 Hyperlane 的鏈聯通。據官方稱，Hyperlane 已支援超過 140 條鏈，網絡累計橋接資產逾 80 億美元。Hyperlane 支援多種虛擬機，並鼓勵用戶「擁有你的跨鏈互通」。by running their own Hyperlane nodes. This architecture offers flexibility and uses open-source bridges (Warp Routes) for token transfers with low slippage. Hyperlane’s focus is on giving developers customizable security modules, hinting that users can choose how trust is managed for each connection.
透過運行自己 Hyperlane 節點。呢種架構帶嚟彈性，並採用開源橋樑（Warp Routes）嚟進行低滑點嘅代幣轉移。Hyperlane 專注於提供俾開發者可自訂嘅安全模組，意味住用戶可以為每條連接選擇點樣管理信任。
DeFi Bridges with Abstraction: Several DeFi protocols have integrated abstraction-like features. For instance, the Synapse and Celer cBridge systems began as liquidity bridges but now offer SDKs and “any-to-any” swaps that hide many details from users. Uniswap’s upcoming UniswapX (leveraging BLOB protocol) and the Across bridge are using “intent-based” designs: users simply request a swap and off-chain solvers execute it optimally. These services blur the lines between a simple bridge and a full abstraction layer, because they often involve relayers and shared liquidity pools.
- 具抽象層嘅 DeFi 橋：有幾個 DeFi 協議已經整合咗類似抽象層嘅功能。例如 Synapse 同 Celer cBridge 系統一開始係流動性橋，但而家已經提供 SDK 同「任意兌換」功能，將好多細節對用戶隱藏。Uniswap 即將推出嘅 UniswapX（利用 BLOB 協議）同 Across 橋，採用「意圖導向」設計：用戶只需要提出兌換請求，鏈下求解器就會以最佳方式執行。呢啲服務模糊咗簡單橋同完整抽象層之間嘅界線，因為佢哋經常涉及中繼者同共享流動性池。
Polkadot and Cosmos (Foundation Projects): While not usually labeled under “chain abstraction,” the architectures of Polkadot and Cosmos historically embody abstraction principles. Polkadot’s parachains share a relay chain and use XCMP messaging so that chains can interoperate without separate bridges. Cosmos’ IBC (Inter-Blockchain Communication) protocol is a built-in standard that lets independent chains send packets of data and tokens to each other. Both systems essentially create ecosystems where chains talk natively instead of via third-party bridges. In Polkadot’s case, developers build on one unified network; in Cosmos, chains install IBC modules to connect. These projects show that abstraction can be achieved at the protocol level (a custom hub chain or native messaging layer), not only at the app level. For example, Cosmos aims to be an “Internet of Blockchains,” where users can send tokens between chains like Terra and Osmosis with IBC without extra steps.
- Polkadot 及 Cosmos（基礎項目）：雖然一般唔會被歸類做「鏈抽象」，但 Polkadot 同 Cosmos 嘅架構其實一直體現咗抽象層原則。Polkadot 嘅平行鏈共享一條中繼鏈，透過 XCMP 訊息令各鏈可以互操作，無需額外橋樑。Cosmos 嘅 IBC（區塊鏈間通訊協議）則係一個內建標準，允許獨立鏈互相傳送資料同代幣。兩者其實都營造咗一個鏈同鏈原生溝通（唔經第三方橋）的生態。以 Polkadot 為例，開發者可以喺同一個聯網部署；Cosmos 就係鏈裝上 IBC 模組就可以連接。呢啲項目證明咗抽象層可以喺協議層（自訂接駁鏈或原生訊息層）實現，唔一定止於應用層。例如 Cosmos 嘅目標係變成「區塊鏈之網絡」，用戶可以用 IBC 輕鬆喺 Terra 同 Osmosis 等鏈之間傳代幣，唔洗再額外操作。

Each of these projects takes a slice of the abstraction puzzle. Some focus on accounts (one wallet for many chains), others on messaging protocols, and others on liquidity routing. But the common theme is that they move cross-chain logic into infrastructure instead of user hands.
每個項目都各自處理部分抽象層問題。有啲專注於賬戶（多鏈一個錢包）、有啲重點放喺訊息協議、亦有啲整合流動性路由。但共同點係，佢哋都係將跨鏈邏輯搬咗去基建層，而唔係用戶手上。

Potential Benefits of Abstraction

Chain abstraction’s promise is multifold. For users, the biggest gain is simplicity. A user logs into a wallet or dApp and no longer needs to know which chain they are using. In the vision of protocols like zkCross (as shared in a recent AMA), chain abstraction makes trading and asset management feel as easy as one-click multi-chain swaps. Users are “freed from complexities like wallets, addresses, and transactions,” so they feel like they are in “a unified blockchain ecosystem”. Instead of juggling ETH gas on Ethereum and BNB on Binance Smart Chain, the abstraction layer might let them pay gas in the token they have, or even in a stablecoin, behind the scenes handling conversions. Interactions like voting, lending, or NFT transfers across chains happen through one interface. All this lowers barriers for non-technical users to try cross-chain Web3.
抽象層帶嚟多重承諾。對用戶而言，最大優勢係簡單化。用戶登入錢包或 dApp 時已經唔需要理會自己用緊邊條鏈。以 zkCross 之類協議（如最近一個 AMA 分享）為例，鏈抽象令交易同資產管理猶如一鍵多鏈兌換咁簡單。用戶「擺脫咗錢包、地址、交易等複雜事」，感覺就似係「身處一個統一區塊鏈生態圈」。唔洗再喺以太坊烧 ETH gas、幣安链又準備 BNB，用抽象層可能直接用任何手頭代幣甚至穩定幣埋單，背後自動搞掂兌換。無論投票、借貸或 NFT 跨鏈傳輸，都可以一個介面搞掂。咁會大大降低唔太熟悉技術用戶試用跨鏈 Web3 門檻。

For developers and protocols, abstraction expands reach. A decentralized exchange or yield aggregator built on an abstraction layer can tap liquidity pools on any connected chain. For instance, a borrowing app could source collateral from many chains and offer a single credit line. Blockworks highlights that developers “can build dApps that are not tied to the liquidity limitations or user base of a specific blockchain,” greatly increasing network effects. This can lead to more capital-efficient applications. In theory, a single Solidity codebase deployed on an abstraction platform might automatically run on Ethereum, Polygon, Avalanche, and more, with the platform handling the environmental differences. This also simplifies devOps – no need to integrate multiple bridge contracts manually. In practice, several abstraction platforms are providing developer SDKs and universal accounts (like Particle Network and zkCross) to enable these cross-chain dApps.
對開發者同協議來說，抽象層代表更大市場範圍。一個建基於抽象層嘅去中心化交易所或收益聚合器，可以任意存取互聯鏈上嘅流動池。例如借貸應用可以跨多鏈攞抵押品，再發一條信用線。Blockworks 指出，開發者可以「整唔受限於某條鏈流動性限制或用戶群嘅 dApps」，增強網絡效應。咁可以帶嚟更高資本效率。理論上，單一份 Solidity 程式碼，只要用抽象平台部署，就自動支援以太坊、Polygon、Avalanche 等多鏈，平台處理執行環境差異。運維上都簡化咗，唔再要手動對接多份橋合約。實際上，已經有啲抽象平台（如 Particle Network、zkCross）提供開發者 SDK 同通用賬戶，方便開發跨鏈 dApp。

Abstraction could also optimize performance and costs. Since these protocols can choose the fastest chain and bundle transactions, they might avoid the congestion and high fees that plague any single network. As Blockworks notes, developers could pick “the most cost-effective and scalable blockchain for specific tasks without regard for their social or economic power”. For example, an NFT marketplace might mint on a low-fee chain while settling final payments on another. Because chain abstraction pools resources, users might get better execution prices (global price pooling) and faster confirmations (parallel settlement). If done well, the system dynamically routes around chokepoints. Some projects even promise cheaper cross-chain transfers via competitive solver bidding.
抽象層都有機會優化執行效能同成本。呢類協議可以揀最快條鏈、打包交易進行，避開傳統高塞車費用問題。Blockworks 都話，開發者可「無需考慮鏈本身聲望或經濟勢力，純粹揀最抵用又擴展性高嘅鏈做某件事」。例如 NFT 市集可以選擇喺低手續費鏈上鑄幣，再用另一條鏈做結算。因為鏈抽象可以共享資源，用戶可能享受到更好執行價（全球價池）同極速確認（平行結算）。如果設計好，系統會動態避開樽頸。有啲項目甚至以競投求解者方式降低跨鏈費用。

Finally, abstraction pushes innovation in UX. Concepts like cross-chain user accounts become real. Imagine logging into a game and having assets from Ethereum, Solana and others all visible at once. Or consider governance: a DAO could have members from any chain vote in one interface. Abstract wallets and signatures are emerging (e.g. NEAR’s “chain signatures” allow one account to sign on multiple blockchains). All of these start to feel like the original promise of Web3 – one decentralized ecosystem rather than siloed silos. If chain abstraction can deliver these, it might greatly accelerate adoption by making crypto tools more familiar and less intimidating.
最後，抽象層推動用戶體驗（UX）創新。跨鏈用戶賬戶之類構想變得現實化。設想登入一隻遊戲，可以一次過見到來自以太坊、Solana 等多鏈資產。又或者 DAO 治理，無論會員由邊條鏈嚟都可統一介面投票。抽象錢包同簽名都已經冒現（例如 NEAR 嘅「鏈簽名」令單一賬戶可以喺多條鏈簽署）。呢一切令 Web3 變返最初諗法 —— 一個去中心化生態，而唔係分割格仔。只要鏈抽象可以真正實現，會令加密工具好似傳統應用一樣親切易用，大大加快入門率。

Challenges and Criticisms

Despite the hype, experts caution that chain abstraction is not a panacea. In fact, it may introduce new forms of fragmentation or risk. One notable voice is Avail co-founder Anurag Arjun, who observes that “most current chain abstraction techniques create even more fragmentation” in crypto. His point is that every abstraction protocol is itself an independent ecosystem with its own security model. Each connected chain has distinct validators or nodes, so unifying them still requires trust across domains. This complexity “is the main bottleneck” in true interoperability.
儘管潮流炙手可熱，專家都警告，鏈抽象唔係萬靈丹，甚至會帶嚟新形式嘅分裂或者風險。舉個例，Avail 聯合創辦人 Anurag Arjun 指出，「目前大部份鏈抽象方案反而令加密領域仲更加分散」。佢意思係，每個抽象協議本身都係獨立生態，各自有唔同安全模型。每條互聯鏈都有自己驗證者或節點，統一管理都仍然要跨域信任。而呢種複雜度「成為真正互操作嘅最大樽頸」。

In practice, a user-facing solution might hide chains, but underneath there are still multiple verification schemes that must coordinate. So far, even promising abstraction projects rely on assumptions (like trusting an oracle network or set of relayers) that not all users may like.
實際上，面向用戶方案可以隱藏鏈名，但底層冇變，仲係要多種驗證方案協調。到目前為止，就算幾好嘅抽象項目，依然要依賴某啲假設（例如信 oracle 網絡、依賴中繼者），唔係每個用戶都受落。

Experts also note that past attempts at interoperability (mostly via bridges) mostly moved liquidity but not governance or application logic, which kept ecosystems siloed. Chain abstraction aims to change that, but it must overcome similar issues. Arjun points out that bridging has been plagued by security issues and “high costs,” and user funds are locked in “siloed” pools under the old model. Abstraction layers still have to solve these problems in new ways. For example, who pays for bridges in an abstraction? If the protocol fronts gas fees, how does it recoup them fairly? Many solutions involve out-of-band payments or revenue sharing with solvers. This economic layer is still nascent and under study.
有專家仲提返，過去跨鏈互操作（大多通過橋）都只係搬咗流動性，但治理同應用邏輯仍舊分家，生態還是分隔。鏈抽象要打破呢個局面，都要解決類似難題。Arjun 指出，橋方案一路以嚟都被安全問題、「高昂成本」困擾，舊模式仲令用戶資金鎖住咗喺隔離池。抽象層要用新方法解決。例如，抽象協議下橋樑成本點算？協議代墊 gas，點公平收番用戶？好多方案要靠鏈外分帳或同求解者分收入。經濟層目前仲未成熟，研究緊。

There is also a danger of centralization. To work efficiently, some abstraction protocols might rely on a relatively small set of validator or relayer nodes (at least initially). If those entities collude or get hacked, cross-chain actions could fail. While providers promise to decentralize these over time, early stages often involve a team or consortium. Critics worry: replacing “bridge validators” with “abstraction validators” may not eliminate trust issues – it could just shift them. And if one abstraction protocol becomes dominant, it could become a new chokepoint (or single point of failure) for Web3 apps.
仲有中心化風險。為提高效率，有啲抽象協議或者要靠少量驗證者或中繼節點（至少初期如此）。如果佢哋同流合污、一齊被 hack，跨鏈操作就會出事。雖則有啲項目話會慢慢去中心化，但初期好多時都係靠團隊或者聯盟管制。批評者擔心：由「橋驗證人」轉做「抽象驗證人」，其實只係信任問題轉移咗，而唔係解決。反之如果最後某個抽象協議一統天下，搞唔掂又變新樽頸或者單點故障。

Another concern is fragmented standards. Several groups are racing to define cross-chain intents and messages (Frontier’s CAKE framework, Ethereum’s ERC-7683, Chainlink CCIP standard, etc.). The market is still deciding on universal protocols. Until there are shared standards or adapters, different abstraction platforms may not interoperate with each other. This could create a new “chain of abstraction” problem: instead of many blockchain silos, we might have many abstraction silos. One developer joked that the era of abstraction might add its own complexity layer. As Mint Ventures puts it, paradoxically “the chain abstraction protocols, which are born to solve fragmentation issues, end up providing fragmented solutions themselves”. In other words, adopting a single abstraction network does not yet eliminate the underlying fragmentation of blockchains; it simply makes it less visible.
另一大隱憂係標準分割。目前好多團隊都爭住推出跨鏈意圖同訊息協議（如 Frontier CAKE 框架、Ethereum ERC-7683、Chainlink CCIP 等等）。市場仲未共識好 universal 協議。無共同標準／適配器之前，各抽象平台其實未必互通。變相有機會出現新版「抽象鏈孤島」：以前係鏈與鏈各自為政，將來可能係抽象網各自為政。有開發者笑言，抽象年代甚至會加一層複雜。Mint Ventures 話得啱，「原本為解決分割而生的鏈抽象協議，自己最後反而製造分割解決方案」。即係話，單靠一個抽象網，暫時都未能消除底層鏈分割，只係遮住咗。

Finally, timing and maturity are factors. Some analysts wonder if chain abstraction is the “next big trend after modularity” or just a hype bubble. Many protocols are in early stages or still testnets. It remains to be seen how fully decentralized and battle-tested they become. Meanwhile, bridges and interoperability issues still exist today, so dApp builders cannot wait indefinitely. For now, developers might use a hybrid: supporting popular bridges while designing for future abstraction networks.
最後一點，時間同成熟度都好關鍵。有分析質疑，鏈抽象到底係「模組化之後新風口」定係炒作泡沫？好多協議依家都係早期或試網階段，仲要觀察到底可以幾去中心化、幾實戰。現階段，傳統橋問題未解決，dApp 建設者唔能夠無限等落去。實務上，開發者多數係混合策略：一面支援熱門橋，另一面為未來抽象網設計。

Are Chain Abstraction Protocols the Future?

The promise of chain abstraction is indeed compelling: a frictionless, unified Web3 where users and apps float seamlessly across blockchains. Some major players are betting on this vision. For example, the team behind ZKCross Network envisions acting “as a central hub, unifying critical elements such as messaging systems and liquidity to enable the creation of easy-to-use cross-chain dApps”. They see chain abstraction as akin to how Visa and Mastercard abstract banking networks in traditional finance. Similarly, projects like Particle Network are already offering unified accounts for millions of users across Ethereum-compatible chains, proving that aspects of abstraction can work at scale.
鏈抽象確實有吸引力：冇摩擦、無縫連接嘅統一 Web3，用戶同應用可以跨鏈自由漂移。有大團隊已經重注呢個願景。例如 ZKCross Network 團隊自比為「中央樞紐，統合訊息系統同流動性，令易用跨鏈 dApp 得以誕生」。佢哋將鏈抽象比做 Visa、Mastercard 等於傳統金融中抽象銀行網絡一樣。仲有類似 Particle Network 等等，已經喺以太坊兼容鏈提供統一賬號俾幾百萬用戶，證明抽象層有機會大規模落地。

However, it’s too early to declare abstraction protocols a panacea. As developers and academics observe, interoperability has many facets. 不過，話抽象協議係萬能靈藥仍然言之過早。開發者同學界都指出，跨鏈互操作仲有好多面向未解決。Early wins have come from both bridges and protocol-level solutions (Polkadot, Cosmos, etc.). It is likely that the future is not one approach exclusively, but a combination. Bridges will continue to connect networks (and they keep getting better secured and optimized), while abstraction layers will grow around them to smooth out UX and liquidity. Over time, standards like IBC or CCIP may interlink many abstraction networks as well, forming a multi-layered web of connectivity.

較早期嘅成功例子，來自橋樑同協議層面嘅解決方案（例如 Polkadot、Cosmos 等）。未來好大機會唔會只用單一方案，而係多種方式一齊用。橋樑會繼續連接唔同網絡，而且會不斷提升安全同優化；而抽象層就會圍繞住橋樑繼續發展，令用戶體驗同流動性更加順暢。隨住時間推進，好似 IBC 或 CCIP 呢類標準都有機會互相串連多個抽象網絡，形成一個多層次嘅連結網絡。

The term “battle of the bridges” suggests competition, but in truth many of these projects complement each other. For instance, Axelar and LayerZero can be thought of as advanced message relays, while Hyperlane and CCIP aim to standardize how those messages look. DeFi apps like Synapse or UniswapX are building user interfaces that will ride on whichever messaging backbone becomes dominant. Meanwhile, engineers continue research on new tech: zero-knowledge proofs to prove cross-chain state, cryptographic light clients in smart contracts, and even blockchain design changes that could eliminate some cross-chain needs.

「橋樑之爭」個名好似講緊競爭，但其實好多項目都係互補。好似 Axelar 同 LayerZero 可以視為先進嘅訊息中繼，而 Hyperlane 同 CCIP 就想將訊息格式標準化。DeFi 應用程式例如 Synapse 或 UniswapX 等，都會以主流訊息骨幹建立用戶界面。與此同時，工程師繼續研究新技術：用零知識證明驗證跨鏈狀態、智能合約入面嘅加密輕客戶端，甚至修改區塊鏈設計，去減少跨鏈需要。

From a user perspective, the coming years should steadily blur chains together. We already see wallets that auto-switch networks, DEXs that source liquidity from multiple chains, and meta-chains like Base or Blast trying to bring projects under one Layer 2 umbrella. Chain abstraction protocols promise to accelerate this convergence by abstracting complexity from the user. If they deliver on their goals, the average crypto user might soon manage assets and use DApps without ever thinking, “Am I on Ethereum or Avalanche right now?” – they’ll just transact, unknowingly spanning chains.

對用戶嚟講，未來幾年唔同鏈之間嘅界線會愈嚟愈模糊。依家已經見到有啲錢包可以自動切換網絡，有啲 DEX 可以喺多條鏈搵流動性，而 Base、Blast 呢啲 meta-chain 就試圖將啲項目拉入同一個 Layer 2 平台。鏈抽象協議聲稱可以透過為用戶隱藏複雜性，加快呢個融合。如果佢哋做到目標，普通加密用戶將來可能操作資產、用 DApp 都唔使再諗自己係用緊 Ethereum 定 Avalanche —— 佢哋只係純粹交易，無意識地跨鏈操作。

Whether that means chain abstraction is the future of Web3 interoperability is still uncertain. The idea is attractive and backed by serious projects and research, but it must prove itself at scale. Observers will watch metrics like total value locked (TVL) in these systems, real-world app adoption, and whether cross-chain hacks truly fall as reliance shifts. For now, the landscape is in flux: bridges remain vital, protocols like Axelar and Chainlink CCIP are rolling out new features, and experiments by projects like UniswapX are testing off-chain solver approaches.

但係鏈抽象會唔會真係成為 Web3 互通性嘅未來，仲未有定案。呢個概念好吸引，亦有嚴肅項目同研究支持，不過最終要喺大規模應用先證明到得唔得。觀察者會留意呢啲系統入面鎖倉總值（TVL）、真實應用採用情況，同埋跨鏈漏洞數字會唔會因為依賴模式轉變而減少。依家個格局仲係轉變緊：橋樑仲係不可或缺， Axelar 同 Chainlink CCIP 呢啲協議都推出新功能，而 UniswapX 呢啲項目都不停測試 off-chain 解決方案。

In conclusion, chain abstraction protocols represent an exciting evolution in blockchain interoperability. They tackle long-standing issues in novel ways and have garnered significant industry attention. Yet they are also one piece of a larger puzzle. The future of Web3 interoperability will likely be layered: a combination of trust-minimized bridges, standardized messaging protocols, modular blockchains (parachains, rollups), and abstraction layers. Over time, these layers may give the illusion of a single Web3. For now, the “battle” is less about winning outright than advancing collectively toward a seamless multi-chain ecosystem.

總結嚟講，鏈抽象協議係區塊鏈互通性一個令人興奮嘅新進化方案。佢用創新方法解決咗長年問題，業界都好關注。不過佢都只係大局入面其中一部分。Web3 未來嘅互通性，好大機會會變成有層次結構：兼容可信橋樑、標準化訊息協議、模組化區塊鏈（例如 parachain、rollup）加埋抽象層。慢慢咁，呢啲層次可能會令人以為淨係得一個 Web3。至少現時，「戰爭」唔係為咗邊個贏，而係一齊推進向無縫多鏈生態發展。