Skip to main content

A Whirlwind Tour of Chain Abstraction

By Thanefield ResearchJuly 5, 2024No Comments


Against the backdrop of Ethereum’s rollup-centric roadmap, the number of blockchains is exploding, with the proliferation of rollups and appchains validating the multichain thesis. Despite its benefits, though, multichain scaling creates a silo effect, with liquidity and users fragmented across multiple environments. This makes it harder than it needs to be for users and developers to use blockchains.

Consider as an example a typical user operating across three blockchains. They face several challenges. First, they must find a secure, affordable, and fast bridge to transfer assets. Using a poorly designed bridge results in overpaying in fees or having assets frozen for long periods during the transfer process.

They also need enough native tokens to pay for transaction fees on each blockchain, which introduces additional friction if the user transacts in non-native assets like stablecoins. Then, there is also the need for them to store and manage each seed phrase separately, again introducing friction and security risk.

Similarly, developers face increased complexity in ensuring interoperability between chains, efficiently managing cross-chain liquidity, or integrating various layers of infrastructure together, all of which significantly increases development time.

Without significant improvements to UX such as those offered by chain abstraction solutions, these complexities are only expected to grow as the number of blockchains and rollups increases.

We believe that chain abstraction is the solution to these challenges. By simplifying user and developer experiences, it can unify fragmented environments and make blockchains more accessible and efficient for billions worldwide. With this in mind, this research piece explores various projects in the abstraction landscape that are helping make this vision a reality.

Chain Abstraction at a High Level

Conceptually, chain abstraction is about simplifying the complexities of interacting with on-chain finance, hiding them from end users and developers.

From a developer-centric viewpoint, the goal is to enable developers to build chain-agnostic applications that operate seamlessly across all rollups quickly and securely, without worrying about the underlying execution complexities.

From a user-centric perspective, the vision is that users will interact with decentralized applications without needing to understand the crypto concepts that power them. It aims to remove all technical complexities and provide an intuitive user experience.

A common analogy for chain abstraction references how we interact with computer applications today. Despite the internet’s ubiquity in daily life, most internet users don’t understand concepts like HTTP, TCP/IP, and the other technical complexities required for it to function. Similarly, when building web applications, most developers aren’t required to have a deep understanding of communication protocols, with browser environments abstracting away most of the work happening at lower layers of the stack like the OS.

For crypto users today though, funds on one chain are inaccessible to applications on another without explicitly bridging. Similarly, a developer’s choice of which blockchain to deploy on still carries outsized weight.

As a result, the current state of crypto is akin to the early days of consumer computing. Chain abstraction will be the driving force that elevates it to the seamless user experience of the modern internet we are accustomed to today.

Eliminating existing UX frictions and simplifying interactions with on-chain applications for both end users and developers will unlock a new wave of growth for crypto. This will drive mass adoption and extend the user base beyond the current siloed web3 native community to billions of users worldwide

Early signs of this vision are evident with Telegram, where 900 million users can be easily onboarded to crypto through a familiar interface. Similarly, Base users can set up Smart Wallets with passkeys, avoiding the need to securely store 12-word seed phrases or pay gas fees, as their transactions can be sponsored by Coinbase.

While both ecosystems are still in their early stages, their progress indicates that this vision is closer to reality than it may appear, and entirely achievable.

Components of Chain Abstraction

Achieving this level of abstraction will require breakthroughs across several layers of infrastructure. Below, we will first break down the building blocks of the abstraction stack. We’ll then dive deeper into each category and highlight important projects working on it and their respective design choices.

Account Abstraction

Account Abstraction (AA) is a concept designed to enhance the user experience for blockchain users by introducing smart contract wallets. As the name suggests, it abstracts away from the end user complexities associated with using blockchain wallets today, such as the need to manage public/private key pairs. AA as a concept emerged within the Ethereum community as far back as 2016, as Ethereum core developers were frustrated with the limitations of existing wallets. It now has its own track in Ethereum’s roadmap, which should culminate with the fully native AA. Although the implementation may vary across different chains, our discussion will focus on Account Abstraction within the Ethereum and EVM contexts.

On most EVM-compatible chains, there are two types of accounts: externally owned accounts (EOAs), and smart contract accounts. EOAs are traditional wallets, such as the ones accessed through Metamask. They are controlled by private keys and used to sign messages and interact with the blockchain. EOAs have several limitations that can significantly impair the average user’s Web3 experience, including the management of private keys, the requirement to pay gas fees in native tokens, and constraints on atomic transactions.

Smart contract wallets are fully programmable and tackle these UX challenges by incorporating Web2 design principles, such as social login systems and account recovery. The method to achieve smart wallet functionality varies depending on the blockchain’s design and the infrastructure developed atop it. In the context of Ethereum and most EVM chains, the network currently does not support native account abstraction, meaning only EOAs can sign messages.

Presently, two standards of smart wallets have gained wide adoption with millions of accounts deployed: Safe, a pioneer in the space, and ERC-4337, a relatively new standard that relies on intents and additional off-chain infrastructure. The upcoming Pectra upgrade is also set to include EIP-7702, which advances the existing account abstraction framework closer to the final stage, where EOAs will be able to transition into Smart Contract Accounts.


As the first entity in account abstraction and the most utilized smart wallet provider, Safe (formerly Gnosis Safe) began its journey as a multisig wallet. It has since evolved into a comprehensive smart wallet solution, becoming a crucial component of the Ethereum and EVM infrastructure. Safe currently has almost 10 million deployed wallets and secures about $90 billion in assets across various EVM chains and rollups.

Safe employs a modular architecture. The core components are integrated into the battle-tested Safe{Core} stack, while Safe Modules introduce custom features, enhancing functionality. This modular approach is akin to the hooks used in Uniswap v4, where Safe Modules ensure robust security at the core level and simplify customization and integration for developers. Developers can create modules to meet specific needs or integrate pre-existing ones. For example, users can add or remove modules that facilitate passkey authentication or manage allowances. Moreover, Safe includes an ERC-4337 module, making it compatible with this newer account abstraction standard and its associated infrastructure.


The current standard on Ethereum and most EVM chains, ERC-4337, was implemented on the Ethereum mainnet in March 2023. It serves as an intermediate step in the development of account abstraction and does not necessitate modifications to the consensus layer protocol for implementation. Instead, it utilizes a concept called pseudo-transactions(user operations), which are based on intents and a combination of on-chain and off-chain infrastructure to facilitate and execute these operations.

ERC-4337 introduces significant enhancements to user experience:

ERC-4337 transaction flow

ERC-4337 introduces a modified transaction flow without altering the consensus layer. This new approach integrates a set of infrastructural components that differentiate it from the typical EOA transaction cycles. The primary differences manifest prior to the transaction being signed, whereas the subsequent process remains unchanged. Key elements introduced include user operations (UserOps), paymasters, alternative mempools (alt mempools), bundlers, and the EntryPoint contract.

In the ERC-4337 transaction cycle, users express an intent to perform a specific action on-chain through a UserOp rather than signing a transaction directly as with EOAs. These intents are managed within the Alt mempool, which is distinct from the public mempool and exclusively handles UserOps. Bundlers, akin to block builders, monitor the alt mempool, selecting UserOps based on the priority fee for inclusion in their bundles. These bundles are then signed by the Bundlers and submitted to the EntryPoint contract, a global contract on Ethereum designated for all ERC-4337 operations, for execution. If necessary, transactions can be sponsored or gas costs can be covered using ERC-20 tokens through the intervention of Paymasters. After these steps, the transaction proceeds in the usual manner and is executed on-chain.

For a visual representation of this process, this diagram provided by Blocknative is highly informative:

Source: BlockNative

The ERC-4337 transaction flow introduces new critical points for MEV extraction. As ERC-4337 wallets generate user operations (or UserOps), there is inherent value in the ordering, exclusion, or inclusion of these intents, just like with transaction in the usual MEV supply chain. Given that the functions of bundlers in ERC-4337 are similar to those of searchers and builders, we think that the most significant effect is going to be an acceleration of the pre-existing trends — more volume will be expressed via intents and through the order flow auctions, circumventing the public mempool.

ERC-4337 adoption

Since its deployment in 2023, ERC-4337 has seen significant adoption on Layer 2 solutions and sidechains, particularly on Base and Polygon. To date, over 5.5 million ERC-4337 wallets have been established, with weekly successful user operations averaging around 800,000.

Coinbase is at the forefront of promoting the development and adoption of smart wallets. On June 5th, Coinbase unveiled the Coinbase Smart Wallet, a new product that incorporates the ERC-4337 standard. This smart wallet offers several notable features, including passkey authentication, sponsored transactions for selected dApps on Base, and the capability for multiple ownership of accounts. With Coinbase’s strategic focus on onboarding new users to the Base platform, it is probable that smart wallets will soon become the predominant wallet type on Base.

Biconomy, Pimlico, and Alchemy have also emerged as leaders in providing essential components of the ERC-4337 infrastructure, most notably in Bundlers and Paymasters. The table below illustrates this dominance in numbers of UserOps executed and paid for.

Despite these encouraging figures, ERC-4337 wallets have yet to achieve widespread adoption on Ethereum mainnet, with only two to three hundred weekly active wallets. Safe wallet remains the main standard for smart wallets on Ethereum. One of the primary limitations of the ERC-4337 design is that it does not allow conversion of the existing EOAs wallets into smart wallets. Additionally, relatively high gas fees on the Ethereum mainnet make some functions, like sponsored transactions, economically unviable.


Following ERC-4337, EIP-7702 represents a significant advancement towards the endgame of full native account abstraction. Drafted recently by Vitalik Buterin, this proposal emerged swiftly as a response to the highly debated EIP-3074, which faced compatibility issues with the future EIPs in the Ethereum AA roadmap. Unlike ERC-4337, which operates at the infrastructure level, EIP-7702 proposes changes directly at the protocol level. It is slated for inclusion in the forthcoming Pectra upgrade, expected between Q4 2024 and Q1 2025.

EIP-7702 arguably represents the most significant improvement proposal tackling user experience in Ethereum’s history. It enhances the ERC-4337 framework by introducing key features such as transaction batching, gas sponsorship, and temporary permissions for EOAs. Specifically, it introduces a new transaction type that allows EOAs to temporarily adopt smart contract code for the duration of a transaction, reverting to their original state upon completion. This proposal ensures forward compatibility with existing ERC-4337 implementations and aligns with the long-term Ethereum AA roadmap.

Case study: Worldcoin

Worldcoin is developing a protocol that they’ve termed proof of personhood, designed to enable apps to verify that users are real persons, not AI-powered bots. This verification is facilitated by World ID, a digital passport issued after a user scans their iris using Orbs, a specialized device. Once obtained, World ID serves as a universal verification tool across various apps and services. In addition to identity verification, users are eligible for bi-weekly WLD grants, which are distributed on-chain.

Worldcoin has successfully issued over 4.5 million World IDs, enabling users to authenticate their personhood without requiring any prior knowledge of blockchain technology. Upon registration, the World App automatically generates a Safe smart wallet for each user on the Optimism network, on the backend. This process completely abstracts the blockchain layer, providing a user experience that mirrors familiar Web2 functionalities such as facial authentication, social recovery, and detailed account management.

Both WLD grants and World IDs are stored in a self-custodial manner, ensuring users retain control over their digital assets. In the case of Worldcoin, Safe-powered smart accounts enable users to access benefits of self-custody and economic incentives offered by crypto while enjoying a Web2-like user experience. The result has been an impressive amount of adoption, onboarding a large number of first-time users into the Web3 space.

Interoperability, Liquidity Aggregation and Intents

With Ethereum’s rollup-centric roadmap and the growing popularity of application-specific chains, the number of distinct blockchain platforms will continue to rise rapidly. This expansion underscores the need for robust cross-chain communication.

Certain ecosystems have developed native interoperability solutions that provide standardized security models and facilitate a degree of chain abstraction within their domains. Notable examples include Polkadot’s shared security architecture and Cosmos’s IBC protocol. In the context of rollups, it is possible to introduce synchronous cross-chain messaging and enable atomic cross-chain interactions by using a shared sequencer, which processes and orders transactions, as well as manages the state. This approach, for instance, has been adopted by Optimism for its Superchain vision.

Despite these advancements, communication across chains, especially outside these established ecosystems, remains a significant challenge due to the absence of native interoperability and widespread standardization. In this section, we will explore various architectural designs in interoperability as they pertain to chain abstraction. Additionally, we will highlight leading projects in each vertical, demonstrating how they contribute to advancing blockchain connectivity.

Message-passing systems

The classic approach to blockchain interoperability utilizes generalized message-passing systems, often reliant on an external set of validators. In this design, a user specifies the desired outcome, and an off-chain entity constructs a precise execution path across two or more chains. This path is executed by a coordinated set of smart contracts and relayers. However, as each chain continually produces blocks and thus changes its state, achieving atomic execution across multiple chains is inherently challenging. Even with a robust data-availability layer that maintains the state of all integrated chains, navigating a route through multiple chains presents significant complexities.

The design choices and architectures of message-passing systems vary widely. They can be modular or monolithic, permissioned or permissionless, support various chains, and operate based on mint-and-burn mechanisms or liquidity pools. Developers tasked with creating a chain-abstraction stack face numerous trade-offs in selecting message-passing systems to integrate, each offering different levels of security guarantees and user experiences. This diversity in design and functionality is likely to hinder the adoption of a universal standard, leading to further fragmentation in the field.

Simple implementations of message-passing systems are currently employed by cross-chain aggregators such as Li.Fi and Socket. These platforms integrate with numerous bridges and DEXes to simulate proposed routes for users. Once a route is selected, it is executed in a strictly ordered sequence.

Intent-based design

In intent-based interoperability design, a more recent development, users simply express a desired outcome rather than specifying an exact execution path as they would in a typical blockchain transaction. These intents are then auctioned off to Solvers—off-chain entities that bid for the right to execute these intentions. The method by which these intents are resolved is secondary; they may be partially or fully matched against each other, or filled from the Solvers’ own inventories. Rather than specifying an execution path, users in this system specify a result, with specialists competing to offer best execution.

A key advantage of this approach, especially in the context of cross-chain asset transfers, is that it deals directly with native tokens instead of IOUs, thereby providing native security guarantees and enhancing overall security. Today, intent-centric applications are primarily found in bridges, such as Across and Synapse, and DEXes including Cow Swap, Uniswap X, and 1inch Fusion.

Recently, Across and Uniswap collaborated to propose a cross-chain intent standard ERC-7683, one of the first attempts at creating a unified framework for intents-based systems to specify cross-chain actions. Other notable developments include Socket’s recent announcement of a new version focused on cross-chain modular order flow, and the announcement by Everclear (previously Connext) of an intent-based primitive leveraging a solver network and an EigenLayer-based Optimistic Rollup to manage liquidity across various blockchains.

However, implementing intent-based solutions poses significant challenges. First, users require access to an interchain account—a smart account that manages keys under the hood and enables transactions across multiple chains. In addition, standardization presents a major hurdle; currently, each intent-centric application must independently develop its entire infrastructure, including intent aggregation, matching, and auction models, with this lack of standardization leading to fragmentation and inefficiency in the broader ecosystem.

Chain abstraction is a concept that doesn’t have a technical specification, and therefore can be approached from many different angles. Some of the most interesting attempts, in our opinion, include Anoma with its intent-centric architecture, Polygon’s aggregation layer, and NEAR’s full-stack chain abstraction solution. We’ll dive into each of them.

Case study: Anoma

Anoma is a privacy-preserving, intent-centric protocol designed for decentralized counterparty discovery, solving, and atomic multi-chain settlement. This platform is unique in its architectural approach: unlike typical blockchain systems where users must specify an execution flow, Anoma requires users to define an end state they are willing to accept, which is expressed through programmatic commitments known as intents. What sets Anoma apart is that these intents are composable and can be collectively resolved, regardless of their origin.

The transaction architecture of Anoma involves the following steps:

  • Generalized Intents: Anoma’s architecture is designed to handle arbitrary intents rather than being confined to application-specific or special-cased scenarios. This flexibility allows for a wide range of potential applications and interactions.
  • Counterparty Discovery: This is a decentralized process in which individual intents are distributed (gossiped) across the network, making them accessible to potential solvers.
  • Solving: In this stage, solvers collaboratively combine and compute intents to find a valid solution—a transaction that can be executed and settled across chains.
  • Settlement: Solutions are verified and finalized on-chain. Anoma’s intent-centric architecture supports settlements on its own sovereign Layer 1 (L1) chain, other L1 chains, or any rollup that settles on an L1.

Source: Anoma

Case study: Polygon AggLayer

Polygon’s AggLayer is a zero-knowledge (ZK) proof-based system that seeks to address the challenges of interoperability and fragmentation across different rollups and Layer 1s (L1s). This approach provides uniform cryptographic security and atomic composability among the connected chains by aggregating ZK proofs from all participating chains.

AggLayer introduces a connected environment to Ethereum through a single unified bridge contract. Each connected chain maintains a copy of this unified bridge root, which enables seamless cross-chain transactions. Additionally, AggLayer features a messaging protocol bridge that establishes Message Queues for each chain, allowing them to maintain local queues of outbound messages secured by ZK proofs. This eliminates the need to lock tokens on one chain to interact with another. By posting ZK proofs of events across multiple chains on Ethereum, AggLayer enables a seamless user experience, akin to interacting within a single ecosystem.

Source: 4Pillars

The Polygon CDK allows projects to launch ZK-based interconnected L2s or connect existing L1s ones to the AggLayer, maintaining liquidity, users, and state. The first components of AggLayer went live in February 2024, marking a significant milestone in Polygon’s roadmap towards creating an aggregated web of sovereign chains.

Case study: NEAR Chain Abstraction Stack

NEAR is developing a comprehensive chain abstraction stack for its blockchain and surrounding ecosystem. The stack consists of the following components:

  • Security Aggregation Stack: This component of the stack includes NEAR DA (Data Availability), which collects states from supported chains. Additionally, it integrates zkWASM, a project being developed in collaboration with Polygon, and leverages EigenLayer’s powered Fast Finality to enhance transaction processing speeds.
  • Account Aggregation: Based on Multi-Party Computation (MPC), this aspect enables NEAR accounts to interact with external blockchains by requesting signature verification. The private keys for these third-party chain accounts are managed by the validators of the NEAR network, functioning as a decentralized signing service. This setup effectively binds accounts across different networks to a central NEAR “master-account,” which can securely manage all associated accounts.
  • Intent Layer: This layer features relayers who execute complex cross-chain intents, facilitating more sophisticated transactions and interactions across the blockchain network.
  • Application Layer: This layer consolidates various web3 services into a single, user-friendly application, streamlining access and interaction with decentralized technologies.

A visual representation of NEAR’s account aggregation architecture is shown below:

Source: Near Balkans Hub

Application Layer

Viewed back-to-front, the application layer is the final stage of Chain Abstraction, where infrastructure is composed and presented coherently to developers and users.

In an ideal end state, developers would be able to easily build chain-agnostic protocols without needing to compose various modular layers, a process that would otherwise present considerable overhead. This involves abstracting away the blockchain selection process, managing cross-chain liquidity, and picking a data availability solution.

From a user’s perspective, an ideal end state would involve interacting with blockchain applications as smoothly as they would with any other digital service, without considering crypto-related points of friction like gas and seed phrases. This involves simplifying user interfaces, streamlining onboarding processes, and eliminating the need for users to understand the underlying technologies, each of which currently presents a significant barrier to entry. Eliminating these points of friction would create a dramatically better user experience and facilitate mass adoption.

Before this vision is realized, tooling to aggregate conflicting infrastructure into a common interface must be built. For that reason, we see chain abstraction as critical to good UX.

Whoever owns the frontend has the most direct relationship the user, and can accordingly capture the most value from their order flow. While the majority of attention and investment so far has been dedicated to infrastructure, we believe that it will soon move up to higher layers of the stack.


There are nearly 300 chains with significant liquidity and on-chain expressiveness, from Layer-1 to Layer-3 solutions. This number is growing, and the trend shows no signs of slowing down.

Among the main drivers for this surge are applications’ demands for scalability and sovereignty, which can be achieved through owning their execution stacks and economics. Examples include recent initiatives by ENS, Aave, and dYdX to launch their own rollups. Open-source technologies like OP Stack have also made it cheaper and easier to build, deploy and operate rollups, with Rollup-as-a-Service providers like Conduit and Caldera further reducing operational and technical overhead. Ironically, deploying a rollup is often cheaper today than transacting on Ethereum during the 2021 cycle.

For today’s users, managing crypto is already often confusing and cumbersome, involving tasks such as securing seed phrases, signing multiple transactions for simple tasks, handling assets across different chains, bridging those assets, and finding the best pricing on various DEXes. While rollups offer the potential to scale without compromising on security and decentralization, their proliferation inherently implies a degree of complexity from the perspective of both users and application developers. Naively implemented, this will only make the UX situation worse.

Modern chain abstraction tooling remedies this, making crypto simpler and more viable to a larger audience. Due to their proximity to the user, the winners in this category will capture significant amounts of value. As on-chain applications generate increasing amounts of revenue, the market will realize the value of owning the frontend.