DEVELOPMENT
Key findings
Contemporary Web3 ecosystems exhibit intricacy, erecting barriers to entry and necessitating users to navigate through a fractured infrastructure. This often results in less-than-ideal user experiences and vulnerabilities, such as the MEV susceptibility.
Intents function as authenticated messages, empowering users to articulate their objectives within the network, while third-party solvers tackle technical complexities, streamlining the process and enhancing user engagement. Intentions pave the way for novel use cases.
Solvers play a pivotal role in uncovering counterparties, discerning and categorizing user intent to optimize valid transaction pathways. Their competitive landscape, rooted in market-driven factors like price and trust, incentivizes them to discover the most efficient execution path, maximizing user advantages.
The account abstraction acts as a gateway to intent, enabled by the verification logic in smart contract wallets. However, its design primarily caters to a singular domain and lacks seamless operability across diverse blockchains.
To address more generalized, multi-domain intents, protocols like Anoma, SUAVE, and Essential are constructing the requisite infrastructure for a specialized intent layer. This involves the introduction of fresh intent languages and virtual machines.
Prominent applications enhance their functionality by directly engaging with solvers. Different protocols handle distinct intent types, such as UniswapX, CoW Swap, and 1inch Fusion.
The heightened composability of intent pools amplifies solver matching capabilities, streamlining the integration of solver networks. Success in intent-driven networks hinges on aligning solver incentives with users’ desired outcomes.
Settings oriented toward intentions present technical intricacies that AI can adeptly alleviate. AI capabilities encompass interpreting natural language intentions, breaking down objectives, and devising optimal transaction routes.
1.Intent-Centric Thesis (Intent-Centric Thesis)
Intentions: what, why, and how
Intentions have recently garnered attention as a remedy for the challenges confronting users within the current Web3 landscape. Despite millions of users engaging in blockchain transactions, a significant gap exists; of the 5.19 billion Internet users, fewer than 100 million possess blockchain wallets, the primary gateways to decentralized applications (“dApps”). What impedes the broader adoption of Web3?
Firstly, navigating modern Web3 systems can prove intricate and time-consuming, particularly for those less versed in such technologies. Steep learning curves significantly elevate entry barriers, dissuading a substantial user base from entering the ecosystem. Interacting with blockchains requires traversing fragmented infrastructures to assemble an execution route. This involves creating and signing transactions in a specific format, potentially across multiple applications and blockchains, and explicitly providing all necessary information for the desired outcome. This complexity often results in suboptimal user experiences (“UX”), is error-prone, and exposes the potential misuse of MEVs by more experienced entities.
As a solution to these issues, intentions have emerged, aiming to simplify user interactions with Web3.
But what precisely are intentions?
Intentions are authenticated messages enabling users to articulate their objectives, with third-party participants, referred to as solvers, handling the technical intricacies to actualize these goals.
Fundamentally, intentions encompass a set of declarative constraints; users delegate transaction creation to solvers, who, on their behalf, determine the optimal computational path while retaining full asset control. From a user’s perspective, intentions permit a focus on ultimate goals—the “what”—without being mired in specific steps—the “how”—to achieve those goals.
Intent-driven solutions can be viewed as an abstraction of order flow, allowing dApps to emulate the UX of conventional applications. Users are liberated from complexities associated with blockchain interaction, entrusted to solvers. This proves particularly advantageous as the industry shifts toward a layered network (“L2”). While Ethereum’s federation-oriented roadmap addresses scalability, it introduces asynchronous states across multiple blockchains, complicating information processing and management between these blockchains. This trade-off adds to the challenges users face in navigating this space. It is envisaged that much of the complexity related to inter-blockchain liquidity, blockchain exchanges, bridge selection, and more is delegated to these server-side solvers.
Ultimately, intentions are crafted to revolutionize the Web3 experience from a manual, step-by-step process to a more intuitive, results-oriented approach. They hold the promise of smoother transactions, swifter execution, and enhanced UX.
Figure 1: Current Web3 order processing is fragmented.
Intentions vs transactions
The current method of conducting transactions in Web3 results in suboptimal user experiences and inefficiencies as users are compelled to make decisions without adequate access to information or sophisticated execution strategies. The shortcomings of transactional inefficiency become apparent when examining the steps necessary for users to execute transactions on two distinct blockchains. From connecting the wallet and switching networks to running the bridge method and signing transactions, this process underscores significant pain points.
In terms of definitions, a transaction represents a precise action a user aims to perform, while intent delineates what the user seeks to achieve within specific parameters. When submitting a transaction, users detail the exact computational path. In contrast, when submitting an intent, users specify the goal and constraints, and a mapping process determines the most suitable computational path. This system grants intentions flexibility, offering multiple execution avenues as opposed to the rigid path set for transactions.
It’s crucial to highlight that the prevailing transaction-based approach also permits users to delegate transactions to a third party, typically applications. However, these applications often lack sufficient incentives to achieve optimal outcomes for users. The innovation of intent lies not merely in delegating transaction creation to a third party but rather to a network of specialized solvers poised to deliver superior results. This arrangement enhances execution efficiency because solvers can assimilate extensive data from multiple blockchains, eliminating the need for repetitive user interactions.
Figure 2: Moving from a transaction-based interaction to an intent-based approach removes additional barriers for users.
The role of solvers
Now, let’s explore the execution of intentions. The pivotal role in this process is played by “solvers.” These solvers are essential as they identify and match intentions with compatible preferences through various intricate methods involving numerous independent counterparts. Upon finding a successful match, these intentions are then settled on the blockchain and undergo verification by the user. This entire procedure is known as counterparty discovery and is carried out by permissionless agents referred to as solvers, employing specialized algorithms. In essence, solvers can be likened to intelligent assistants operating within the blockchain realm.
Solvers, acting as intermediaries, bridge the gap between user intentions and blockchain execution by engaging in competition to offer the most favorable prices for execution. The solver that submits the most appealing execution price earns the right to execute the user’s order. As solvers vie for superiority in both price and reliability within a free market, this competition serves as an incentive for them to identify the optimal execution path and deliver maximum benefits to users.
Solvers employ various channels to match intentions:
Liquidity provision: Solvers can function as the primary counterparty, offering their liquidity for transactions.
Partial filling: Multiple independent parties collaboratively purchase the token being traded.
Direct Coincidence of Wishes (“CoWs”): Two intentions may directly clash with each other.
Ringing transactions: Intentions can be resolved without direct CoWs. For instance, three intents can create a balanced transaction even if no pair of intents satisfies each other’s preferences.
Solvers often take on specialized roles and participate in diverse groups. For instance:
Solver DAO: These DAOs act as generalized solvers, potentially facilitating cross-ecosystem interactions. They may consist of solvers specializing in specific Tier 1 (“L1”) ecosystems and engage in private solutions using trusted execution environments.
Application-specific solvers: Focused on particular industries or verticals, these solvers specialize in specific applications, refining their strategies for tasks on DeFi platforms or gaming applications.
Funds solver: Specialized organizations that may focus solely on routing or intent matching solutions without maintaining any inventory, while others may act as market makers and simultaneously hold inventory, playing a dual role.
Individual solvers: Users or communities can sometimes address their own intents by connecting to intent networks, managing solvers on nodes embedded in local devices.
In summary, the solver network is a crucial element of an intent-centric roadmap that utilizes off-grid infrastructure to enhance the user experience within the blockchain.
Figure 3: Solvers act as coordinators tasked with identifying user intentions, categorizing them, and designing the best path to create valid transactions.
Relevance of Account Abstraction
While the concept of account abstraction (“AA”) is not within the immediate focus of this document, ERC-4337 introduces a feature for wallets to function as a gateway to intent. This functionality is made possible by the embedded verification logic in smart contract wallets. AA enhances external accounts (“EOAs”) by enabling them to be managed through smart contract wallets or by allowing smart contracts to directly initiate transactions. Users primarily relying on EOA can delegate their existing transaction-based workflows to solvers.
As a result, EA is tailored to fulfill “special intentions,” often described as “gasless transactions” and “seedless recovery.” It reduces entry barriers for personal wallets and offers users more user-friendly account systems. Although still in the early stages of implementation, the initial market reception, as depicted in Figure 4, has shown promise. However, realizing its full potential requires further innovation and development.
Intentions simplify user objectives, and account abstraction streamlines the process of achieving those objectives.
A notable drawback of AA is its exclusive focus on user accounts, designed primarily for use within a single domain. It lacks the capability to seamlessly operate across multiple blockchains or support cross-network payments. This limitation arises from the fact that ERC-4337, while capable of providing a specific set of intents within the Ethereum framework, is not a generalized system for broader intent applications. For instance, AA still requires users to manually determine the most efficient route for order flows such as swaps, bridges, or liquidity provision. Intent aims to address this issue by allowing users to specify only the start and desired end states, removing the need for this level of detection.
Addressing UX issues with AA alone will not completely eliminate the array of problems users currently face. However, the recent surge in AA adoption indicates a market demand for products that enhance user experience. With AA capabilities, intent is expected to drive further adoption.
Ultimately, the true appeal of AA lies in its architecture, which transforms wallets into an entry point for intent. Nevertheless, its limitations become evident in multi-domain scenarios—a challenge that the intent framework aims to overcome. For more in-depth information on AA, refer to our recent report: Tutorial on Account Abstraction.
Figure 4: Month-on-month increase in ERC-4337 smart account transactions signifies wider adoption, highlighting the potential demand for appropriate intent-driven products.
2.Market Landscape
The fundamental concept of intentions has a longstanding history. Even before the advent of Uniswap and automated market makers (“AMMs”), various protocols already incorporated intent-based order books. Additionally, NFT marketplaces have utilized signed intentions for listings and bids for several years. Another noteworthy example is found in decentralized exchange (“DEX”) aggregators, where users focus solely on achieving optimal execution, remaining indifferent to specific details such as the choice of DEX.
More recently, platforms like CoW Swap and UniswapX have emerged, presenting more advanced solutions for limit orders based on intent.
In the current market landscape, there are varying perceptions of intentions; some liken them to traditional transactions (“txs”), while others see them as a contemporary form of limit orders. However, the architectural design for these intentions is typically straightforward and single-purpose. Intentions encompass much more than these classifications imply, with numerous innovations yet to surface.
While many systems have emerged to cater to the limit order option, the development of more advanced intent-based tools paves the way for broader general-purpose architectures. Notable examples include Flashbots’ Anoma and SUAVE, which function as general-purpose systems introducing an intent layer. Users can broadcast signed intentions to Gossip Nodes in these systems.
Intent-specific blockchains within these systems aim to bridge the gap between users signing intentions and solvers executing them across networks. Ultimately, once the infrastructure is established, these general-purpose solvers are designed to offer more robust use cases for intent-specific applications.
Use cases
Intentions represent a valuable tool for enhancing counterparty selection, automation, and the coordination of multi-lateral engagements. Presently, the industry is witnessing the emergence of several intent-specific applications, each unlocking a diverse range of use cases. Here are a few noteworthy examples:
1. Limit orders and batch auctions: Limit orders serve as partial transactions, with resolvers competing to identify the optimal combination of counterparties. This may span multiple decentralized exchange (DEX) pools, ensuring optimal routing and pricing for users.
2. Smart orders: Users can specify the execution terms of their swaps and utilize intentions to place custom rates tailored to their requirements.
3. Automated Actions: Automation can be applied to execute various actions, such as dollar-cost averaging (DCA) in a token or automatically rebalancing portfolios within a set range or at predetermined times.
4. Cross-bridging: Bridging processes can pose a significant user experience challenge. By specifying preferences and risk thresholds, users can delegate the process, enabling solvers to manage it efficiently.
5. Crowdfunding: Through intentions, users can contribute funds under specific conditions, committing only when a project reaches predetermined milestones. An example is GitCoin Matching Funds, where users donate in advance to winning projects before their selection.
6. Peer-to-Peer (P2P) networking: Users can transact directly with others by expressing their intent, eliminating the need for intermediaries and seeking better prices. In lending contexts, lenders can determine terms, such as desired collateral types and rates. If a borrower repays early and the lender wishes to continue lending, solvers can match them with a suitable counterparty.
7. Security Verification: Intent can be used to restrict the interaction of smart contracts that offer verifiable evidence, such as proof of approval by whitelisted audit groups.
Ecosystem Mapping
The visual representation presented below offers a broad perspective on initiatives delving into the realm of intentions. It is crucial to acknowledge that there might be certain intersections between categories, and thus, the diagram is presented in a simplified manner for clarity. Notably, the DeFi sector witnesses a proliferation of use cases driven by intent. Anticipated in the future are substantial shifts, extending beyond the DeFi sector to influence other sub-sectors.
Figure 5. Visual map of the different intent-driven projects and their categories.
3.Project overview
Before we delve deeper, it’s essential to underscore that the intent sector can be broadly categorized into two segments: general-purpose infrastructure and intent-specific applications.
General Purpose: Intentions of a more general nature require an architecture tailored to optimize their utilization. Various protocols are advancing this infrastructure by introducing components like new intent languages and virtual machines (“VMs”).
Intent Specific: Decentralized applications leading the charge in specific types of intents extend their functionality by directly involving solvers.
With these distinctions, different protocols cater to specialized intent types. For instance, platforms like UniswapX and CoW Swap implement exchange-oriented intents. Intents focused on a single domain or wallet can be addressed by AA wallets or decentralized applications compatible with Essential. On the other hand, SUAVE and Anoma position themselves to manage generalized, multi-domain tasks.
In the context of intent-driven applications, while solvers play a pivotal role, the absence of solutions with generalized intent poses challenges.
One notable challenge is the necessity to maintain separate infrastructures for solving and executing off-blockchain. While some teams ambitiously tackle this challenge, these efforts typically involve multi-year plans that can impede application-level innovation. Operating within custom, isolated off-chain networks inadvertently results in the protocols forfeiting the ability to compose at the intent level.
Platforms like CoW Swap and 1inch Fusion, operating in the DEX aggregator space, operate with separate and non-overlapping intent pools. This fragmentation hampers solvers’ potential aggregation and matching capabilities, impeding the propagation of network effects crucial for the fundamental growth of these ecosystems.
Intent-driven networks thrive when solvers competitively resolve intentions, driven by the accumulation of commissions. The more intentions are shared, the greater the incentive for new solvers to join, essential for improving the quality of intent fulfillment. This underscores the critical need for a dedicated intent layer with the requisite infrastructure. Leveraging general-purpose solutions such as Anoma is designed to alleviate this burden and redirect application attention to product-related issues. However, the dynamic between general-purpose and purpose-driven solutions is likely to evolve in the future.
Figure 6. Comparative overview of general-purpose and intent-specific solutions.
General purpose
Anoma
Anoma stands out as a singular intent-driven network with a vision to construct a global intent network where nodes can observe and process intent across a diverse range of use cases.
Efficiency and confidentiality in counterparty discovery are facilitated by Anoma through the utilization of intentions. Anoma specifically provides declarative privacy, decentralized counterparty discovery, decision processes, and multi-blockchain atomic settlement.
Unlike positioning itself as a function within a specific stack or layer, Anoma positions itself as a foundational architecture deployable at any layer (L1, L2, etc.). It is crafted to enable the creation of fractal instances of protocols by anyone, collectively forming the Anoma ecosystem. This approach offers significant advantages in terms of flexibility and the ability to compose each fractal instance.
Within the Anoma framework, users articulate desired outcomes through intentions, prompting the protocol to generate signed partial transactions. These intentions are then transmitted to intent gossip nodes, creating permissionless intent pools functioning as a P2P network layer filled with diverse user intentions. Resolvers navigate these pools to discover counterparties with matching intentions, collaboratively creating transactions to fulfill the intent conditions of each user. Interoperability of resolvers allows them to work across all fractal instances, facilitating intent matching across multiple blockchains.
Anoma’s gossip layer employs path authentication to enhance efficiency, particularly in complex intent scenarios requiring a series of mappings. Intermediate solvers and gossip nodes are tracked and rewarded for their participation. When a solver partially fulfills an intention, it can claim rights to its efforts and share the partial intention with other solvers. Rewards are then shared between matching intentions based on path authentication, fostering continuous collaboration and helping solvers maintain local trust graphs.
A crucial aspect of Anoma lies in its modular, open-source architecture that developers can leverage when creating new applications. The technical stack of Anoma comprises various components, giving developers the flexibility to use the entire stack or select specific elements to address particular use cases, such as counterparty discovery and problem-solving.
Figure 7: Users can send intentions to the network, with Anoma performing both counterparty discovery and computation.
Figure 8. Intentions are propagated through Anoma’s intention propagation nodes, while player matching nodes use solver algorithms to identify and match compatible intentions.
Namada
Namada serves as the inaugural fractal instance within the Anoma framework, focusing on settlement-level privacy. It functions as a testing ground for Anoma’s secure multi-asset pool (“MASP”). Anoma’s MASP employs zk-SNARK to enable all assets to share a common secure pool, enhancing user anonymity. Consequently, the owners of these assets can form a large anonymous group, collectively bolstering each other’s privacy.
The primary objective of Namada is to ensure compatibility with Ethereum or any IBC-enabled blockchain. It achieves this by supporting both fungible and non-fungible tokens transmitted through a dedicated Ethereum bridge or IBC. Privacy-preserving asset transfers to Namada leverage MASP, enabling assets like NFTs, ETH, DAI, or any of Namada’s proprietary assets to utilize the same secure set. This approach renders them indistinguishable from external surveillance. The protocol incentivizes users to retain protected assets, as they contribute liquidity, aligning with the intentions represented by the users.
Typhon
Typhon functions as the core system responsible for storing, organizing, and executing transactions within Anoma, playing a crucial role in areas pertaining to consensus, gossip, parallel execution, and mempools. Anoma’s consensus mechanism is grounded in Heterogeneous Paxos, specifically designed to facilitate atomic inter-blockchain transactions across all fractal instances. In simpler terms, it empowers Anoma to fulfill intent concurrently across multiple blockchains. The foundational components are the underlying blockchains, representing fractal instances of Anoma, and chimera chains facilitating atomic transactions between these underlying blockchains.
This setup enables multiple underlying blockchains to connect into a chimera chain, fostering seamless interactions.
While atomic inter-blockchain transactions typically involve multiple rounds of consensus between blockchains, resulting in increased latency, Anoma introduces parallel execution of unbound states. This innovation reduces the necessity for multiple rounds of consensus, subsequently mitigating latency issues. This capability opens avenues for features like inter-network flash credits, atomic swaps, and other advanced functionalities. Anoma further employs a Narwhal-based mempool for efficient transaction sequencing. In essence, Typhon strives to accelerate transactions within the blockchain, streamline complexity, and enhance the ability of applications to interact across multiple overlapping instances.
Taiga
Taiga operates as an execution environment, delivering tools for privacy-preserving applications by enabling componentized privacy. It facilitates privacy in the context of Account Abstraction (AA) by empowering users to define their privacy settings. Intents within Taiga can be classified into three categories: public, screened, and private. While users have control over their privacy settings, it’s important to note that solvers, in interpreting intents for their functionalities, may expose certain elements that aren’t entirely private.
Applications built on the Taiga framework store their state as annotations attributed to the application, incorporating logic encapsulated through validity predicates, akin to smart contracts. These validity predicates effectively function as a form of Account Abstraction, establishing the authorization logic for the application.
Ferveo
MEV emerges from the transparency of transactions, making intentions vulnerable to extraction when disclosed before execution. Ferveo, a pivotal component of Anoma, addresses this issue by employing distributed key generation (“DKG”) and threshold encryption. This approach enables the creation of private mempools until the transaction order is confirmed.
DKG operates by generating a single public key from distributed segments of the private key, known as public keys. Users submit their intentions, signing them with this public key, ensuring their encryption until the transaction order is confirmed. Upon confirmation, validators leverage threshold signatures to decrypt the transactions. Notably, the decryption process requires the consensus of two-thirds of the majority validators, providing robust crypto-economic privacy guarantees.
In essence, Anoma’s implementation of an intent-based architecture stands out for placing user-centered design at the core of the protocol. This lays a robust foundation for the continued development of other protocols. While the full implications of intentions are yet to be unveiled, Anoma is anticipated to play a pivotal role in shaping their future implementation.
Figure 9: Anoma uses DKG and threshold encryption to encrypt user intent until the transaction order is validated.
SUAVE
SUAVE by Flashbots, which stands for Single Unifying Auction for Value Expression, strives to enhance user value while promoting decentralization in public blockchains.
Functioning independently of the blockchain, SUAVE acts as an autonomous intermediary layer to facilitate intent. It introduces MEVM, a variant of Solidity featuring new precompilations tailored for MEV use cases. This adaptation empowers developers to construct MEV applications as smart contracts within an expressive, familiar, and flexible programming environment, akin to the conventional EVM.
By distinctively separating the roles of mempool and block builder from existing blockchain structures, SUAVE introduces a highly specialized, decentralized, off-the-shelf alternative. Its architectural framework revolves around three key elements: a mempool where users express their preferences, an execution network where solvers vie to fulfill those preferences, and a block-building environment where solvers create blocks for adoption by other networks.
SUAVE, through the utilization of a common consistency layer, fosters decentralization, enables block developers to tap into cross-domain MEV, allows validators to optimize revenue, and ensures users superior transaction performance. This configuration also serves to counteract the centralizing impact of MEVs. At its core, SUAVE revolves around preferences—messages signed by users to express their intentions, facilitating both simple transfers and complex sequences across multiple blockchains. In essence, solvers compete for optimal execution by capturing MEVs and offering decentralized order flow value in the process. Overall, SUAVE aspires to become a universal mempool and blockchain builder for all blockchains.
Figure 10: SUAVE separates the roles of mempool and block builder from existing blockchains, offering a highly specialized “plug and play” alternative.
Essential
Essential has a dedicated focus on establishing a comprehensive suite of solutions within the intent architecture domain, centering around three primary objectives: developing a subject-specific language (“DSL”) for expressing intent, creating an Ethereum standard for intent-oriented Account Abstraction (AA), and constructing a modular intent layer.
DSL: Essential adopts a specialized DSL constructed in Rust for expressing intent. This DSL serves as the groundwork for a standardized expression of intent, optimized for solvers, thereby enhancing layout and contributing to the evolution of intent-based applications.
Intention-based AA standard: The intention-based AA standard empowers solvers to formulate valid transactions based on user intent.
Modular Intent Layer: The modular intent layer establishes the groundwork for an exclusive architecture, consolidated order flow, resilience to MEV, and the capability to execute intent between blockchains. In essence, it functions as Essential’s protocol for intent processing.
Within Essential’s architecture, intentions undergo resolution rather than direct implementation. This resolution generates an execution trace of the resolved intent for execution on the blockchain. Diverging from other solutions, Essential opts against using cryptographic execution to ensure privacy. Instead, the underlying intent is recorded on the blockchain before decryption and execution. Furthermore, Essential’s consensus mechanism incentivizes solvers to compete for objective satisfaction, fostering competition for high satisfaction among end users.
In summary, a unified standard mitigates fragmentation issues stemming from various intent types. This streamlines the collaboration between solvers and intent-enabled applications, facilitates developers in implementing intent systems, and eliminates the need to replicate similar infrastructure multiple times.
Intent Specific Intents
UniswapX
The emergence of UniswapX in July of this year marked a significant shift in the industry towards intent-specific applications, representing a strong endorsement of the intent-centric approach within DeFi’s well-known decentralized exchanges (DEXs).
The platform operates as follows: users articulate their intentions, which are then processed off-blockchain by specialized implementers—whether solvers, market makers, or search engines—before being executed on the blockchain. This stands in contrast to Uniswap’s Automated Market Maker (AMM) model, where users directly interact with AMM pools on the blockchain. With UniswapX, users circumvent AMM entirely by directing their transactions to executors who employ both on-net and off-net strategies to competitively fulfill the transaction.
UniswapX leverages Dutch auctions to handle intentions, commencing with a high price that gradually decreases until the bidder finds it favorable and completes the order. This structure, in a competitive market, aids in reducing slippage, fostering a more favorable environment for order flow auctions.
Participants have a range of options for executing swaps, including through Uniswap, Balancer, or Curve pools, aggregators, or utilizing their own inventory while hedging on centralized platforms.
A notable feature is that users can interact with the Uniswap front-end and process transactions using fully off-chain liquidity sourced from centralized exchanges (CEX). This positioning somewhat establishes Uniswap as a gateway to CEX liquidity.
Beyond conventional user experience enhancements, UniswapX brings additional advantages in cross-chain transactions. Users are no longer required to undergo the multi-step operations necessary for exchanging assets between blockchains. Additionally, traditional interchain bridge designs often involve funds being held in a vulnerable bridge contract, susceptible to hacking. However, with UniswapX, the only funds at risk are “transit swaps” or active exchanges, markedly reducing the exposure during the exchange process and enhancing security, especially in light of the recent recognition of September 2023 as a significant month for crypto exploits.
The introduction of Uniswap V4’s hooks, allowing easy integration of capabilities with existing liquidity pools, underscores the focus on enhancing the online trading experience. UniswapX streamlines interconnectivity, optimizing efficiency and user experience. Ultimately, this effort to improve the online trading experience plays a crucial role in bridging the gap between decentralized and centralized exchange trading volumes, with the intent-centric approach serving as a key factor in achieving this equilibrium.
Figure 11. Swappers generate subscribed orders specifying their exchange’s targets, and executors utilize competitive strategies to satisfy those orders.
Figure 12. The daily number of transactions on UniswapX is trending upward.
CoW Swap
Having amassed a cumulative trading volume exceeding $28 billion to date, CoW Swap stands as a pioneering force in the realm of intent-based trading applications, introducing peer-to-peer settlement orders through the innovative CoW protocol. While the trading volume is notably high compared to peers, the platform’s full potential may not have been realized yet, likely owing to its recent launch.
The protocol employs batch auctions, where a competitive market of solvers vies to secure the most optimal batch through individual strategies. CoW Swap incentivizes solvers to match buy and sell orders by bundling them together for execution at the best possible price. Solvers in CoW Swap face a clear dichotomy: they either possess the best batch or they do not, with no middle ground. They can execute these orders offline against each other, leveraging CoWs—instances where two or more traders directly exchange assets without relying on in-network liquidity, eliminating the need for liquidity provider fees.
Resolvers in CoW Swap exhibit flexibility in their strategy, tapping into various sources of liquidity, including liquidity pools on the blockchain, term loans, or even private order flows, although the latter is suboptimal. If additional liquidity is required, the rest of the transaction can be handled by identifying the most efficient route through liquidity pools on the blockchain. Unlike platforms with dedicated liquidity providers, CoW Swap integrates with existing liquidity on the blockchain and can also utilize liquidity from other AMMs if necessary. Currently, CoW Swap can process orders on platforms like Uniswap, Sushiswap, 1inch, and Paraswap, as well as execute custom transactions via signed messages.
CoW Swap extends its capabilities to multi-dimensional swaps involving three or more asset transactions. While these trades don’t directly offset each other, the idea is that liquidity is shared across multiple trades, achieving an overall balance when executed together. This addresses concerns about fragmented liquidity, allowing the protocol to match multiple token pairs in a single batch and settle at a single price, thereby saving on gas fees.
Despite the benefits of using intent demonstrated by CoWs, CoW Swap encounters challenges as it operates in an ecosystem not inherently designed for intent. Ethereum’s structure doesn’t fully align with the intent structure, leading CoW Swap to establish and maintain a separate off-network structure. This limited scope affects only the intent of its direct users, reducing access to wider intent mempools and diminishing the frequency of CoWs in CoW Swap.
Nevertheless, CoW Swap’s journey with intentions is evident in its early days, primarily seeing batches won by DEX aggregators like 1inch, facilitated by CoW Swap’s integrated 1inch API. Over time, the emergence of specialized solvers such as Barter, Otex, and Laertes has diminished the role of 1inch and other aggregators, showcasing the growing effectiveness of solvers in maximizing value and optimizing batch orders. Additionally, CoW Swap recently introduced CoW Hooks, allowing users to perform DeFi actions with custom code, enhancing layout efficiency and abstracting complexity for users.
Figure 13. Although trading volume exhibits volatility, weekly CoW Swap values typically exceed $200 million.
Figure 14. Several overlapping intentions reduce the need for AMM on the blockchain, as evidenced by this CoW, where 12 swaps efficiently distribute liquidity.
Figure 15. The popularity of solvers on CoW Swap is growing, while the market share of DEX aggregators is decreasing
1inch Fusion
In response to the rise of intentions, 1inch took a distinctive approach with the introduction of 1inch Fusion, offering an alternative avenue for order matching outside the CoW Swap network. While CoW Swap utilizes batch auctions, 1inch Fusion employs Dutch auctions, aligning its operational mechanics more closely with UniswapX. Within the Fusion ecosystem, users, acting as creators, dispatch orders or intentions off the blockchain. These intentions are then made accessible to solvers engaged in the Dutch auction facilitated by 1inch.
1inch aids users in automatically generating an intent based on the prices provided by their DEX aggregator. However, users maintain the flexibility to tailor or choose different order parameters. Similar to CoW Swap, the direct matching of intent can circumvent the necessity for traditional liquidity pools.
Since its inception in December 2022, Fusion has witnessed substantial sales volume, capitalizing on its integration with the 1inch protocol and its sizable user base. The shared networks of solvers between 1inch Fusion and CoW Swap demonstrate a natural and cost-effective overlap within the ecosystem. As the emphasis on intent-driven solvers intensifies, one can anticipate robust competition among solvers across various protocols and networks.
Figure 16. Since its inception, Fusion has maintained stable monthly transaction volumes, currently contributing an average of over $1 billion in transactions per 1inch.
4.Prospects
Opportunities and Risks
In the realm of decentralized products, a significant hindrance to widespread adoption has been the often challenging user experience (UX), favoring centralized alternatives. However, innovative concepts like SUAVE and Anoma are spearheading a transformative shift, placing user intent at the forefront of decentralized activities such as swaps and bets. By prioritizing outcomes over processes, intentions bring forth several capabilities, outlined below.
1. Improved UX: Intentions streamline the user experience by eliminating the need for users to navigate complex transaction parameters. Rather than grappling with a fragmented Web3 infrastructure, users can simply express their desires, letting solvers handle the rest. This user-centric approach has the potential to drive broader adoption.
2. Reduced MEV Impact: Intentions abstract away specific transaction details, reducing the likelihood of anticipatory development and other MEV extraction strategies. This abstraction fosters a fairer transactional environment.
3. Increased Competitiveness: The competition among solvers encourages them to offer the most favorable execution prices, enabling users to optimize profits and minimize slippage.
4. Greater Compatibility: User intentions open avenues for enhanced system compatibility. Solvers can be employed to extend functionality, emulate atomicity in blockchains, and improve overall compatibility among different protocols or dApps.
5. Gas Efficiency: Intents can be consolidated into a single transaction, reducing gas fees and enhancing the cost-effectiveness of transactions compared to multiple separate transactions.
6. Improved Privacy and Security: Expressing ultimate goals through intentions eliminates the need to map out each intermediate step, enhancing transaction security and reducing the impact of user strategies or preferences.
While intentions offer significant benefits, challenges persist, including issues related to centralization, privacy, and resilience to MEV strategies. Balancing solver efficiency with user privacy is crucial, as sharing more information can compromise user privacy. The risk of centralization, where a few dominant players handle complex user intentions, is another concern that needs careful consideration. Additionally, overcoming technical barriers, such as understanding user desires and navigating the blockchain landscape, is essential for the successful implementation of an intent-driven model. Achieving the right balance between efficiency, decentralization, and privacy remains a key consideration in the evolving landscape of intention-driven systems.
Neighboring Considerations
Providing adequate incentives for solvers
Challenges confront intent-driven platforms in effectively engaging solvers to meet user needs. In the short term, the constraints of low volumes may limit the optimal structuring of incentives, potentially diminishing solver effectiveness. However, with the increasing momentum in the field, it is anticipated that heightened competition will organically lead to the reallocation of solver incentives, aligning them more closely with user requirements.
In all intent-based systems, solvers require incentives to cover operating costs and generate profits, often operating as profit-oriented entities with a propensity for short-term strategies. In their pursuit of these incentives, solvers engage in competition with other solvers and generic MEV bots, establishing themselves as integral components of the MEV ecosystem. Optimal solver incentive schemes seek to align the incentives of solvers with the ideal outcomes desired by users, ensuring that increased MEV for solvers translates into faster and improved execution for users. For instance, a solver’s commission might be determined based on the tokens involved in the user’s intent. Allowing solvers to claim a portion of fees and share their partial solutions within the overall framework adds further utility. Therefore, the incorporation of robust incentive structures for solvers constitutes a pivotal aspect of intent-based architectures.
To foster synergies between users and solvers, the introduction of reputation-based systems could provide additional incentives. These systems would cultivate user trust through consistent interactions and the establishment of accountability. As users delegate search tasks to solvers, the latter become accountable and build a reputation for trustworthiness over time. Users, in turn, can delegate the adherence to blockchain rules to proposers or validators, creating a structured and trustworthy framework within which solvers operate. This integration of reputation-based systems enhances the overall incentive landscape, reinforcing the collaborative relationship between users and solvers in intent-driven architectures.
Potential interaction with AI
The potential integration of advanced artificial intelligence could have a positive impact on an intent-driven approach. Intention-oriented settings encompass various technical aspects, and AI has the potential to simplify these complexities. Whether it involves interpreting intent in natural language, breaking down goals, or determining the optimal route for executing transactions, AI serves as a powerful tool. However, in situations where user intent is complex or ambiguously expressed, solvers may struggle to comprehend and devise optimal solutions. The analysis of user requests and transaction data by AI can assist in bridging this gap.
Nevertheless, the incorporation of AI from conception to implementation is likely to involve multiple parties, raising security concerns. Providers of intent-driven protocols must establish deterrents and penalties for malicious behavior to maintain a secure third-party execution layer. Furthermore, enhancing technical security to prevent vulnerabilities is crucial for safeguarding user rights. While the combination of AI with intent-driven protocols holds promise, it remains an open area of research.
5.Final thoughts
As the adoption of Web3 technology gains momentum, it becomes crucial to empower users to navigate the complexities of the ecosystem independently. In this evolving landscape, intent-driven architectures are emerging as a promising solution, indicating a progressive shift in Web3 interactions. While intentions are still in the research phase, they hold the potential to enhance Web3 user experience and pave the way for the next wave of user engagement.
Fundamentally, intentions strive to abstract users as much as possible, all while preserving decentralization and enhancing the efficiency and expressiveness of transactions. The core objective is to improve functional interoperability to a level where overcoming complexities between different blockchain ecosystems becomes a thing of the past for users. Transitioning from an imperative to a declarative paradigm allows users to articulate their intentions, leaving the intricacies of execution to solvers. Solver networks, coupled with interoperability mechanisms like AA and bridges, will play a pivotal role in bringing practical applications of intentions to fruition.
Several intent-based applications are already in existence, and this trend is poised to continue. As the blockchain ecosystem evolves, the widespread adoption of intent has the potential to revolutionize the way users interact with decentralized applications in the future.
