Foreword

Web3's cumbersome user experience has become one of the reasons that hinder the large-scale adoption of blockchain. With the recent introduction of the 10 potential trends that the well-known crypto venture capital firm Paradigm focuses on in the encryption field, "Intent-Centric protocols and infrastructure" ranked first, and this concept instantly attracted industry attention. This is a design concept that revolves around user needs and views intent as the core. For example, "I want to order a 30 yuan hamburger takeaway" is an "intent". To complete this intent, users only need to enter their name, phone number, delivery address, and place an order on the takeaway platform, without worrying about how the 30 yuan they paid is earned by the merchant or how the platform allocates the delivery rider and how the rider delivers it to their home. This architecture greatly simplifies the user experience threshold, allowing each user to simply express their intent, hand over the intermediate process to various other protocols, and wait for the final result. Note: This article is not for project promotion and is suitable for general readers to read. The content is for analysis and reference only.

Intent-Centric Design Principles

Background:

In the Web3 world, transactions are the most core function, and no matter the DeFi, GameFi, NFT or any other track, they cannot do without basic transactions. However, the reality is that the decentralized nature of blockchain determines that different chains are independent islands, and we need to find bridges that link them to complete asset exchanges. Although centralized transactions provide convenient user asset trading experience, users still need to authorize wallets one by one to enjoy countless Dapp applications. In order to reduce the experience threshold of decentralized applications, the concept of Intent-Centric has emerged.

Although the prototype of "intent" did not originate recently, the concept storm was triggered by Paradigm, so we borrow their definition as a reference: "Intent is a set of signed declarative constraints that allow users to create transactions for third-party outsourcing without giving up complete control over the transaction party." In practice, the signature is dominated by the user's spontaneous intention, but can transferring assets to third-party outsourcing really reduce the user's usage threshold? The following will discuss what content is involved in realizing the vision of Intent-Centric.

What is Intent:

Intent-Centric can be understood as "centered around intent". In the article "Intent-Based Architectures and Their Risks" published on June 1, 23, the concept of intent is explained. In the user's regular transaction process, the transaction signature allows the verifier to verify and execute according to a specific calculation path, and the Gas fee in the process is used to incentivize the verifier to complete the calculation. However, intent does not specifically refer to a particular calculation path, and any path can be used to achieve the final result under specific constraints. In practice, by signing and sharing intent, users actually allow recipients to choose the transaction execution path on their behalf (as shown in the figure below). In order to distinguish it from transactions, we define intent as an information signature, which allows state transitions from a given initial state to retrieve multiple paths to achieve the final state result [1].

Source: Intent-Based Architectures and Their Risks - Paradigm

 As shown in the above figure, when submitting a transaction, a clear execution path needs to be specified. For example, when exchanging tokens on Uniswap, the execution path is determined by the matching result when submitting the intention. For users, they only need to express their intentions, and the related protocols/applications of Intent-Centric will complete the rest, that is, matching the execution path. The execution operation is briefly described as follows: the Solver is responsible for matching the execution path that meets the intention, and finally waiting for the final execution and obtaining the result.

In short: Trading = I specify the execution method to do this; Intent = I only want this result but do not care about the implementation process. Different preferences of intent can be matched in complex ways, and the process of achieving this requires "discovering trading partners" and "solver solving".

Bob，the Intent Solver

Source: ETH Global

 Bob the Solver (pronounced as "Bao Bo Qiu Jie Qi" in Chinese, hereinafter referred to as "the Solver" but not specifically Bob Solver) is an intent-based transaction infrastructure. Its main purpose is to simplify the operation process, reduce the user's participation threshold, and integrate into wallets and decentralized applications. Essentially, it creates and executes outsourced transactions for users. Its implementation is mainly composed of two related parts:

● Abstract Account Wallet for executing transactions: The AA wallet system, which adopts the EIP-4337 standard, consists of a bundler and a paymaster contract, and is integrated with a solver to simplify the transaction process.

Intent Layer and Solver

Can a separate intent layer be built and allow solvers to compete for solutions? Although it is theoretically possible, the reality is that it is very difficult to implement. To address the core MEV problem, Anoma and SUAVE (details can be found in the following content) are mentioned here. In order to achieve the goal of Intent-Centric, what features are required to safely complete asset transfers by packaging transactions for third-party processing?

1. Trustworthy

In order to achieve trustworthy and decentralized blockchain features, it may not be enough to have only a simple "intent network". This is because it cannot fulfill the guarantees of users and solvers for intent. After users indicate their intent, they will be calculated by Sovler for the specific costs required for each execution path, and Sovler will also be constrained by the adjustments set by users. Users need to know that each execution path is trustworthy.

2. PrivacyThe solver provides users with an optimal execution path, but the complexity involved inevitably makes it vulnerable to hacking attempts aimed at stealing assets or obtaining user information. Additionally, in most cases, a user's on-chain activity information is exposed to observers. To ensure the security of user information, it is necessary to encrypt or hide at least some important information. However, achieving the privatization of computable information on the blockchain is extremely difficult.

3. Expression of Intent

Intention is an abstract concept. The way users express their intentions on the blockchain seems to have similar issues to the way users ask questions to AI. How should I explain it to make it understand what I want to express? Some expressions may contain implicit intentions, and the solver may incur more gas fees when choosing the execution path. An efficient and accurate parser is the key to implementing intentions.

4. Maintain Consistency between Intent and MEV

Take SUAVE's independent blockchain that solves EVM real-world problems as an example. When it processes Ethereum intentions, it needs to perform cross-chain settlement. Similar to the huge success achieved by MEV decentralized solutions such as MEV-Geth and MEV-Boost, it meets the needs of cross-chain MEV with a more fair and transparent transaction processing mechanism.

5. Anti-censorshipAccording to Paradigm's explanation, we can easily find that there is a core issue. An AI-enabled parser should not exist as a single entity. If a single parser is attacked or disabled, it will lead to a catastrophic system crash. Additionally, there are issues such as whether the parser exists, refusal to execute, and incorrect execution. These issues may be resolved by Anoma mentioned in the following text.

6. Competitiveness of solvers

a. Permissionless Intentpools: Using an open design that allows anyone to submit intentions to the memory pool and provides unrestricted access to executors without permission.

b. Permissioned Intentpools: User intents that require permission to submit and execute, allowing users to pass intents to trusted third parties to execute on their behalf.

c. Hybrid Solutions: Combining the characteristics of the above two memory pools, aiming to achieve a balance between openness and controllability.

The implementation of the intention layer requires account abstraction (as explained in the SUAVE section below). EIP-4337 consists of six parts: UserOperation, Bundler, EntryPoint, Wallet Factory, Paymaster, and Signature Aggregator. The brief operational process is as follows:

● The entrance contract will pre-send the transaction to the wallet contract or the payment contract.

● The payment contract can be expanded into multiple types based on the project's business logic, and will not be elaborated here for brevity's sake. 【2】.

● UniswapX: off-Chain extension plan intent

● MPC Wallet: Using private key sharding technology, the private key is divided into multiple parts and stored on multiple independent nodes. When performing transactions and other operations, these nodes jointly calculate and generate signatures through secure multiparty computation protocols, without the need to reconstruct the complete private key.

Regardless of representative projects or concepts, they are all simplified for users through a series of more complex instructions. If we divide the above content, it can be classified into four categories:

Based on centralized resource matching platforms, such as Unibot for robot trading on CEX exchanges and Friend.tech, etc.

● Intelligence intent:

Source: Researcher@tmel0211

MEV

Maximal Extractable Value (MEV) refers to the additional revenue rewards that miners can obtain by adding, deleting, and rearranging transactions. This behavior can be used for DEX arbitrage, front-running, back-running, liquidation, sniping bots, time-bandit attacks, sandwich attacks, and other activities that harm the rights of users who use DeFi normally. To elaborate, transactions on Ethereum are sorted by miners based on Gas, with high Gas being prioritized for packaging and processing, and low Gas being processed slowly. Transactions are first submitted to the Mempool, where they wait to be included in a block. Validators extract transactions from the Mempool and add them to the next block when constructing it. Because the Mempool is public, searchers have the opportunity to pay validators fees to order transactions in a specific way and extract value from users through sorting, thus creating MEV miner value. To implement user intent, transactions need to be handed over to third-party outsourcing for processing, so MEV generated by transactions in the Intent-Centric architecture is also one of the key issues that need to be focused on.

The most direct impact brought by MEV: while damaging the entire network, competition also makes the market more efficient.

● Unequal distribution of capital (large holders owning more ETH chips) may lead to centralization of validators (larger staking pools obtaining higher MEV returns), which reduces the overall security of the network. Even though there are some mitigating measures in place, the centralization risk brought by block construction rights cannot be completely eliminated at present.● To increase the likelihood of transactions being packaged, MEV competing searchers compete for priority by bidding for Gas, which can cause the public memory pool to become congested due to high Gas fee transactions from searchers. However, DEX arbitrage and loan settlement can also help the DeFi market reach equilibrium faster to maintain market stability.

MEV is a topic that has been explored for years in the encryption industry and cannot be avoided. How to alleviate the adverse effects of MEV is also constantly being explored. This article will be explained in the next section, "A Universal Solution for Achieving Intent." According to EigenPhi data as of September 15, 2023, there is still a large profit margin for on-chain MEV, and the use of Flashbots' block-producing nodes alone has earned over 200,000 ETH after the Ethereum merger. Therefore, this is a very large profit distribution cake.

Source: EigenPhi | MEV Data

Cross-chain, Sorter and Oracle

Intent-Centric is a large-scale system architecture. According to Paradigm's explanation, the content of Intent-Centric will involve the entire blockchain field, and the massive asset transfers between various Layer1 and different Layer2 need to be passed to the intent for convenient processing. Nowadays, the development of the industry has brought us into a multi-chain era, where each chain is an isolated island, but there are different bridges between the islands. Therefore, cross-chain and sorter are also necessary bridges to meet the intent.

Based on the prosperity of the current Ethereum Layer2 ecosystem, the four recognized giants in the industry, Arbitrum, Optimism, zkSync, and Starkware, are showcasing their Layer3 solutions, while OPstack and ZKstack provide corresponding narratives. The common problem faced by Layer2 currently is the centralized sequencer problem. Although feasible solutions are actively being sought, the reality is that there is a huge cake hidden here. Taking OPstack as an example, we can simply understand its profit model as "renting shops". The profit of Layer2 = Layer2's gas income + MEV income - Layer1's gas expenditure. In order to achieve the intended purpose, the toll fees between Layer2 and Layer1 are essential.

The topic comes to cross-chain bridges. The prosperity of Layer2 has made us aware of the considerable profits of the sorter early on, but the profits of cross-chain cannot be underestimated. Cross-chain is an essential bridge to connect different chain islands. As one of the infrastructure, in addition to meeting the cross-chain needs of normal trading users in the current bear market environment, the cross-chain needs generated by the airdrop expectations of the "Lumaodang" are still actual market demands. The continuously prosperous Dapps applications will also gradually erode profits from convenient centralized exchanges. However, the security of cross-chain bridges is worth noting. According to official data from hacked.slowmist, in the two years from June 29, 2021, to September 16, 2023, there have been 38 public security incidents in the cross-chain bridge ecosystem, with a cumulative financial damage of up to 2 billion US dollars. In summary, the gas fees involved in achieving the intention are inevitable, but this article does not discuss the security of the sorter and cross-chain, although this is a problem that deserves our attention.

Source: hacked.slowmist

 The topic comes to the oracle, and from a macro perspective, the strong financial attributes of blockchain make it a system environment that pursues certainty. Even though the narrative of RWA has been discussed for many years, the reality is that blockchain cannot obtain real-world data outside the chain and can only obtain on-chain data. This is because the virtual machine (VM) cannot allow smart contracts to have network calls, so the operation of smart contracts must result in consistent outcomes. Therefore, for the outside world, the data on the blockchain is also closed.

We zoom in to a microscopic perspective, and the oracle is an important factor in the DeFi world. Although the security of different protocols is usually inherited from the underlying smart contract network, its normal operation still depends on the oracle. If an oracle of a protocol is attacked or destroyed, the entire protocol will be manipulated. Nowadays, DeFi hopes to define itself as "primitives" and hopes that more teams will build products or combine protocols on their basis. However, the new DeFi contracts derived from this iterative process will upgrade their operating logic to support a larger ecosystem, which also brings some external dependencies and unpredictable risks.

After several years of development, the DeFi industry has also accumulated billions of dollars in theft, such as in March of this year, Euler Finance's lending protocol was hacked, resulting in a loss of up to $200 million. It allows users to post collateral and borrow, and its problem occurred in a specific function without security checks, allowing users to disrupt the basic invariant of the lending market (such as the basic invariant in UniSwap: tokenBalanceX * tokenBalanceY == k). Similarly, derivative protocols that rely on oracles for pricing and lack internal price discovery mechanisms are susceptible to price lag and lack of updates, severely limiting their scale and user experience. This also explains why trader Avraham Eisenberg was able to successfully attack Mango Markets and withdraw $116 million from the cryptocurrency trading platform.

In summary, the dark rules of the blockchain forest are generally full of unknown risks, and there is still a long way to go to achieve the vision of the Intent-Centric architecture.

Source: Chainalysis

Aggregator and Gas

The direct purpose of the aggregator is to save users the trouble of finding the best trading path and profit strategy, which includes many types of aggregators, such as trading aggregators, information aggregators, profit aggregators, liquidity aggregators, and asset management aggregators. The Intent-Centric architecture requires a certain degree of centralization, and various types of aggregators may provide a "reference answer" to the execution path of the parser in the process of finding intent.

Gas is an indispensable toll fee for executing various transactions, and how to optimize Gas fees is also a perennial topic in the industry. Currently, the overall optimization is mainly focused on product end, account abstract wallet, and DeFi protocols. In the future, whether the combination of aggregator, account abstraction, and DeFi protocols can achieve the effect of optimizing Gas from a brand new product end perspective remains to be seen.

Wallet Authorization

The first step in Dapps interaction: wallet authorization. Let's focus on the transaction itself again. The purpose of the Intent-Centric architecture is to simplify transactions and reduce user barriers, but each transaction in the intent involves countless authorization signatures. How to solve the authorization problem safely and conveniently is also a problem that needs to be considered. Perhaps account abstraction and dappOS V2 mentioned later in this article are good solutions.

Universal Solution for Implementing Intentions

Anoma

The content you provided is:

Source: Anoma Official

Introduction:

According to a report by Coindesk on May 31, 23, Anoma Foundation has successfully completed a total investment of $25 million from 14 capital firms including CMCC Global and Electric Capital. The architecture of the project's intention was proposed in the white paper released in August 22, which includes the Intent Gossip Layer for spreading intentions, discovering transaction opponents, and matching execution paths. Its design is the default validation path, and all gossip information is signed by the sender node, forming a signature chain that can be traced back to the initiator. This feature is particularly important in resisting censorship and Dos attacks, so Anoma is essential when discussing the implementation of Intent-Centric.

Anoma was initially developed based on Tendermint and used the Byzantine Fault Tolerance (BFT) consensus mechanism. Later, it switched to another PoS proof-of-stake consensus mechanism called "Typhon" developed by Heliax to allow for consensus partitioning between independent chains. The project team has developed various technologies such as Taiga (a private state transition framework integrated into the Intent Gossip Layer and Matching Layer), Typhon (a cross-chain atomic transaction consensus mechanism), MASP (a multi-asset shielding pool), Vamp-IR (an arithmetic circuit language), and Juvix (a smart contract programming language) to promote experiments in cryptography and distributed systems.

Continuing from the previous paragraph on "Programmability of Intent", to achieve the vision of an Intent-Centric architecture, programmable algorithmic structures and the ability to discover counterparties are essential features. This needs to be built on a foundation of multi-interactivity and multi-application, which leads us to introduce the unified architecture of Anoma for full-stack decentralized applications, as an example for building decentralized applications together.

Source: Anoma

Design architecture and its innovative points: 

● Privacy Payment 

In order to protect user privacy and prevent others from tracing and collecting data, the sender, receiver, amount, and asset denomination are all encrypted. The transfer of funds is ensured through zero-knowledge proof zk-SNARKs. Its unique feature is that it allows all assets to share the same shielded pool (MASP) to provide composable asset protection and increase user anonymity sets, rather than shielding each asset separately. The more participants and assets involved in the transfer, the more frequent the transfer, which greatly increases the degree of anonymity of asset transfer data.

Source: Delphi Digital - Delphi Creative

● Barter Trading

● Intent Matching System 

In Ethereum EVM, transactions do not require future states but authorize specific execution paths, while Anoma includes a matching system that allows users to broadcast transaction intentions using Gossip. Simply put, Anoma consists of two main parts: a distributed ledger and an intent matching system, which complement each other and can also run independently. By using Intent Gossip to run nodes, Token exchange solvers, and RPC servers requesting new intentions, transactions are submitted from matched intentions to the distributed ledger, helping users automatically discover transaction counterparts.

● Multi-chain support

Anoma uses the cross-chain communication IBC protocol of the Cosmos ecosystem for blockchain communication. IBC utilizes relayers to achieve data transmission between different blockchains, with the aim of becoming a multi-chain privacy layer. Although relayers are typically operated by node operators at present, anyone with the ability to do so can run them and earn fees in the process.

● Fractal scaling expansion plan

Anoma uses fractal solutions to solve the scalability problem of blockchain, allowing users to create local instances to meet other transaction needs. Fractal refers to dividing Anoma into different application chains to handle different tasks, allowing each Anoma application chain to be highly customized to achieve scalability and cope with user growth (similar to supporting the IBC protocol to achieve expansion goals). In the future, its security will be improved by the ⅣInterchain Security, ⅣMesh Security, and ⅣInterchain Alliance projects.

SUAVE

Introduction:

According to the news on July 25th, 23, Flashbots, an Ethereum infrastructure service, completed a $60 million Series B financing with a valuation of $1 billion to develop the SUAVE platform. SUAVE (Single Unified Auction for Value Expression) separates the memory pool and block generation from the existing blockchain to form an independent layer of blockchain network (sorting layer), and provides highly specialized plug-and-play alternatives, aiming to solve the infrastructure challenges related to MEV.

Although SUAVE is a new blockchain, it is not a general-purpose smart contract platform that can compete with Ethereum or any other participating chains. If Anoma is compared to building an intent-centered blockchain, then SUAVE is building an intent-centered infrastructure.

Source: SUAVE in the blockchain stack

SUAVE's three goals:

● Illuminate the Dark Forest: Expose unfair events in the opaque MEV ecosystem to every user, quantify their impact, and break down information barriers between participants.● Democratic mining: integrates miners and retrievers in an open manner, promotes competition among all parties to the maximum extent, and is open to all miners for free to prevent MEV integration and centralization.

● Profit distribution: MEV is given to the creators of it - Ethereum users.

 According to the SUAVE solution, cross-chain settlement is required when processing external transactions (such as Ethereum) to attempt to solve various risks in the EVM, similar to the huge success achieved by MEV decentralized solutions such as MEV-Geth and MEV-Boost. However, it should be noted that users will deposit funds into SUAVE during the implementation process of the intention, and in the process of unlocking funds when the demand is resolved, the smart contract on SUAVE will need to verify assets and involve oracles, so to some extent, SUAVE will also face all the risks of cross-chain bridges today.

From SCW and AA to SUAVE and Intent-centric:

Looking back at the "Account Abstraction AA" section mentioned earlier, and considering the main functions brought by account abstraction: private key recovery, gas-free payment, multi-signature authorization, one transaction for multiple transfers, fee rate restrictions, etc. Overall, account abstraction perfectly combines the advantageous functions of regular accounts (EOA) and smart contract accounts (CA).

The vision of SUAVE is to become a common sorting layer between different types of chains (including cross-chain transactions and MEV). Therefore, when user intentions involve cross-chain asset transfers, the Account Abstraction (AA) and multi-signature smart contract accounts (SCW) such as Gonsis Safe can combine their strengths and, together with SUAVE's EVM solution, may be the currently theoretically optimal solution.

CoWSwap

Source: CoW Protocol

Introduction:

If you want to select the project that is closest to the Intent-Centric architecture, CoWSwap is definitely on the list. Unlike other protocols, users only need to send a signed order to delegate the execution of the transaction to the resolver network and complete the transaction in that network (without specifying the execution path). At the same time, the off-chain signed order will be executed after being matched by the resolver (similar to the Bob Solver mentioned above, but different). Due to the resolver's "batch bundling" right, the gas fee during the process will be borne by the resolver, and there is no need to pay if the transaction fails.

CoWSwap Hooks：

Main function:

● Create LP position: You can use only pre-hook trading or only post-hook trading;

● JIT (Just-In-Time) Smart Order: Use the composable CoW framework (ERC-1271) to program Safe smart contract wallets to execute custom smart contract approvals and exchanges.

● NFT: NFTs can be sold or purchased within the CoW Hooks feature.

● Cross-chain: Assets can be sent to the bridging contract through Cow Hooks to complete cross-chain transactions;

● Unlocking and re-staking: If you have staked 32 Ethereum as a validator, you can unbind it to claim rewards, or re-stake assets on other chains through CoW Hooks.● Automatic increase of LP position: Assuming that you become a liquidity provider in the EUR-USDC pool on the Gnosis chain to receive GNO Token rewards, you can use CoW Hooks to automatically convert your GNO to EUR-USDC at a 50/50 ratio through CoW Swap, thereby increasing your LP position automatically.

dappOS V2

Whether it is for the purpose of intention or transaction, the user's focus is always on the total amount of funds in the account, rather than the balance of individual assets between different chains. According to the understanding of dappOS V2, there should be a unified account wallet to achieve the vision of "one signature completes everything", and the design concept of one-click multiple "TXN (Transaction)" is introduced to remove barriers for users before completing TXN, and achieve transaction goals guided by user intention.

For example, borrowing it to achieve interaction with GMX between different chains in Arbitrum and Avalanch without assets, so as to achieve the intention-centered basic transaction rules for future DeFi. Therefore, realizing the concept of intention-centeredness requires involvement of chain abstraction, account abstraction, and more "protocol abstraction". However, when it comes to this, one needs to consider a question: assuming that all kinds of abstractions can be implemented, will the interaction rules for users and the airdrop rules for various projects change accordingly?

Summary:

Intent is still full of risks and challenges in its execution process. It relies on intermediaries or specific executors, which may lead to the risk of centralized power and monopoly, thus affecting the trust issue of the entire middleware. Secondly, the security and privacy risks associated with transferring transactions to third-party execution should also be considered. It is highly anticipated that Intent application developers can achieve a balance between security, privacy, and convenience.

The concept of intent has been maturely applied in the Web2 field, such as taxi-hailing software, ticketing software, and map navigation. However, the success of these applications is based on the sound foundation of Web2 infrastructure. To have mature application scenarios in the Web3 field, we still need to wait for the continuous maturity of the industry. Intent-Centric is a grand concept that will involve all aspects of blockchain and is also one of the best paths for combining blockchain with AI. However, whether it can be implemented and developed and flourish still requires our continuous attention.

Explanation of literature and reference articles