🧞 EigenLayer wants to change intents. How?

Fixing Intents

Jan 20, 2025

Gm and happy Monday frens🖖

Imagine you want to send tokens from one of your wallets to another. But the second wallet is on a different chain that doesn't have a native version of your token. So you have to use a third party bridge and a wrapped version of the token. Sounds like hell, doesn't it?

Intents are created to combat this terrible UX - you just say what result you need to get and the solver does it all for you (basically).

However, there are some problems with this. Fortunately, there are also some solutions:

🎉 Events that you must see next month
🏋 Making intent great again
🧵 DEFAI, innovating science & predicting future
📰 News

🔮 Your plans for the February

We've got 4 events next month!

In Hong Kong we'll be hosting our next Telegram Builders Day (Feb 17th) with topics like user acquisition for Telegram apps, successful viral case studies and how to build successful Telegram games!

And in Denver we'll be hosting 3 events:

🥩 Restaking & DeFAI Day (Feb 25th) - expect a day packed with keynotes, workshops, and panels diving deep into the latest in the restaking space.

🧩 Modular & L2 Day (Feb 26th) - side event for web3 developers, tech founders, and enthusiasts seeking in-depth of modular stack and layer 2 scaling solutions.

👾 Agents Day – AI x Web3 (Feb 25th) - where visionaries, innovators, and pioneers will gather to explore the powerful convergence of AI and Web3. Together, we will push the boundaries of what’s possible in building a decentralized, transparent, and equitable future for AGI.

🛠️ Deep-dive: Fixing Intents

How to make modular intent stack scalable, interoperable, and decentralized intent-based ecosystems that can unlock the full potential of dApps?

Session by Ishaan 0x (Protocol Researcher @ EigenLayer) at Chain Abstraction & Intents Day in Bangkok. Watch the full version here.

Current limitations of intents

Intents in their current form, as seen in platforms like CowSwap and UniswapX, suffer from significant fragmentation. While these systems work well individually, they operate in isolation, preventing interoperability across platforms. For example, CowSwap stores user intents on a centralized server that only its whitelisted solvers can access. This reliance on centralized storage creates two major weaknesses: a centralized storage point where intents reside and a centralized selection process where the platform determines which solver fulfills the intent.

This setup presents scalability challenges. As the ecosystem evolves, envisioning a future where every smart contract is intent-based, isolated systems with their own separate "mempools" will be unable to scale effectively.

The solvers themselves remain siloed, limiting their ability to operate across multiple intent-based applications. This lack of interoperability stifles innovation and the broader adoption of unified solutions in Web3, restricting the potential growth of the ecosystem.

Challenges with existing approaches

The reliance on centralized servers and solvers introduces vulnerabilities and inefficiencies. Platforms like CowSwap exemplify this by running proprietary auctions to select solvers, reducing transparency and leaving the system open to censorship risks.

In response, some projects have tried using Ethereum's mempool to propagate intents. While this approach avoids centralized storage, it brings new challenges. Gas costs make it expensive to broadcast intents, particularly for those that might not be executed.

Furthermore, the mempool is vulnerable to spam and DoS attacks, as bad actors can flood it with meaningless intents, overwhelming nodes and validators.

A significant barrier to adoption is the inability to offer a gasless experience. Gas fees discourage users from broadcasting intents, yet removing gas introduces the risk of spam and abuse.

Additionally, scalability concerns arise because the number of intents is expected to far exceed actual transactions, much like canceled limit orders in traditional financial systems.

Proposed modular intent stack

To address these limitations, Ishaan proposed a modular intent stack composed of three critical components: representation, propagation, and selection.

Representation focuses on standardizing intents across applications. ERC-7683 provides a framework for defining cross-chain intents, specifying source and destination chains, token types, and amounts.

This standardization simplifies solver operations by creating a common language for intents, enabling solvers to work seamlessly across platforms. It acts as an "ERC-20 for intents," fostering composability and paving the way for a unified intent ecosystem.

Propagation, aims to decentralize the storage and distribution of intents. Decentralized data availability layers are proposed as a solution to replace centralized servers and the Ethereum mempool.

A DA layer, such as EigenDA, offers censorship resistance, verifiability, and horizontal scalability. These layers allow intents to live and propagate efficiently, supporting the high volume of intents anticipated in the future. By providing a reliable infrastructure, DA layers ensure that intents remain accessible and secure without centralized control.

Selection, involves decentralizing the process of solver auctions. Instead of relying on centralized platforms to run auctions, a leaderless auction model inspired by Sorella Labs is proposed.

In this model, solvers access the same information and collectively determine the best solution for an intent. This decentralized approach enhances transparency, trust, and efficiency, supporting both modular and composable intent ecosystems.

Key Design Principles

The modular intent stack is built on two fundamental principles: verifiability and censorship resistance.

Verifiability ensures that every component of the system operates as intended, fostering trust and integrity in the decentralized infrastructure.

Censorship resistance aligns with blockchain's core values, ensuring that all intents can be stored, propagated, and executed without interference.

Together, these principles create a robust foundation for the next generation of intent-based applications.