Written by: KarenZ, Foresight News
In Web3, issues such as blockchain ecosystem fragmentation, low capital efficiency, and lack of automated trust still exist. Against this backdrop, the Newton Protocol is committed to building a verifiable automated system for the on-chain economy.
Newton enables users to delegate agents to execute on-chain operations through verifiable automation (combining TEE Trusted Execution Environments and Zero-Knowledge Proofs ZKP), while ensuring that actions conform to preset rules. This article will explore Newton’s introduction, team and financing background, operational mechanisms, and how to participate from four aspects.
What is the Newton Protocol? What is the background?
The Newton Protocol is an infrastructure jointly built by Magic and the Magic Newton Foundation, aiming to introduce a verifiable automation layer for the on-chain economy. Newton ensures that each agent's operations are executed within boundaries set by users and utilizes encrypted verification by combining Trusted Execution Environments (TEEs) with Zero-Knowledge Proofs (ZKPs).
Magic Labs, behind Newton, was founded in May 2018, eliminating the complexities of seed phrases and browser plugins through embedded wallets, enabling ordinary users to easily enter the Web3 world.
Magic Labs was co-founded by Sean Li and Jaemin Jin, engineers from the University of Waterloo in Canada.
Sean Li: Serving as the CEO of Magic, he co-founded Kitematic (later acquired by Docker and developed into Docker Desktop, which is used by millions of developers monthly).
Jaemin Jin: Serving as the Chief Product Officer of Magic, he worked as a software engineer at Uber from September 2015 to June 2018, and also spent 10 months as a software engineer on the Apple Siri team.
Magic states that its team members are talents from technology and crypto companies such as Coinbase, OpenSea, and Alchemy.
Since its establishment, Magic has focused on embedded wallets and API wallets, with its SDK capable of being integrated into any web or mobile application, helping to achieve passwordless authentication, non-custodial wallet-as-a-service, customizable embedded wallet widget UI, and compatibility with over 30 blockchains. As of April 2025, Magic reported over 50 million wallets integrated and support for applications like Polymarket, WalletConnect, Helium, and Immutable.
In terms of financing, Magic has raised approximately $90 million in total.
In July 2021, Magic completed a $27 million Series A financing round led by Northzone, with participation from Tiger Global, Placeholder, SV Angel, Digital Currency Group, CoinFund, and Cherubic. Angel investors include Reddit co-founder Alexis Ohanian, former Coinbase CTO Balaji Srinivasan, former Adidas VP Ben Pruess, Vercel CEO Guillermo Rauch, GitHub CTO Jason Warner, and AngelList co-founder Naval Ravikant.
At the end of May 2023, Magic announced the completion of a $52 million strategic financing round, led by PayPal Ventures, with participation from Cherubic, Synchrony, KX, Northzone, and Volt Capital.
How does Newton operate?
The operation process of the Newton Protocol involves participants such as users, developers, operators, and verifiers, as well as components like smart accounts, zkPermissions, and execution coordinators.
The following is the operational process of Newton:
1. Developers build automation agents and package them as containerized applications: Using the Newton SDK and zkML framework, define the logic, constraints, and interfaces for agents to interact with users and protocols, converting their off-chain services into verifiable agents.
2. User initiates requests: Users submit automation intention orders.
3. Order matching and task allocation: The execution coordinator matches users' automation intentions with operators through a limit order book mechanism, selecting suitable operators to execute tasks.
4. Operator task execution: After accepting a task, operators execute defined automation logic off-chain, utilizing TEEs (Trusted Execution Environments) to ensure trustworthy computation, or running zero-knowledge proof circuit machine learning model agents, embedding verifiable execution directly into Zero-Knowledge Proofs (ZKPs). After completing the task, operators will generate and submit TEE proofs and ZKPs.
5. Result verification and execution: TEE certification and ZKP are verified on-chain; if approved, the approved actions will be executed through limited session keys, completing the tasks and updating related states. If not approved, the tasks will be rejected, and operators may face economic penalties.
6. Reputation system updates: Update the reputation of operators based on execution history, task success rates, and user feedback. Well-performing operators will earn a better reputation, attracting more user orders, while underperforming operators may face economic losses and declining reputations.
According to Newton, its three core components are:
1. Smart accounts: Using standards like ERC - 4337, EIP - 7702, ensure the security of user wallets and support fine-grained and revocable permission authorization in an encrypted manner. Users can securely delegate operational permissions to automation agents through policy-based delegation without giving up their private keys.
2. zkPermissions: A zero-knowledge proof permissions system, each session key is bound to a zero-knowledge circuit zkPermissions, capable of encoding automation rules, covering core permission types such as:
Data-based execution conditions: Execute trades only when on-chain sentiment data is bullish; in addition, advanced trading strategies like golden crosses, slippage, liquidity conditions, and Gas settings can be configured, and even scenarios where the inflation rate is less than 2% after the CPI data is released can be set.
Risk checks: For example, volatility-gated execution, setting oversold buy executions, and price difference-triggered limit orders.
Trade volume, frequency, and time limits: For example, setting the maximum number of trades per day, and the upper limit on the volume of each trade.
Enforce baseline session key permissions for each agent, including common permission types such as spending limits, expiration dates, and token allowance lists.
3. Execution coordinator: As a decentralized automated task marketplace, it matches user intentions with operators and verifies the correctness of executions.
In summary, the Newton Protocol provides a new infrastructure for the integration of AI agents and DeFi by transforming automation into verifiable on-chain primitives, combining the efficiency of TEEs and the verifiability of ZKPs, promoting the development of a programmable economy and a user-driven on-chain ecosystem.
How to interact?
1. Earn points on the portal at https://www.magicnewton.com/portal/rewards, ways to earn include:
Roll the dice;
Clear the minefield game;
Social tasks (side quests).
2. Register for the Newton App (https://newton.xyz/app) and complete agency tasks. It should be noted that currently, it is not recommended to deposit a large amount of funds, as the Newton App only supports fund inflows and does not support withdrawals.
3. Post high-quality tweets related to the project. The Newton Foundation will allocate 0.75% of its token supply to the Kaito community as rewards, with 0.5% allocated to quality yapping and successful referrals, and 0.25% distributed based on community milestone achievements to a broader Kaito community (stakers, holders, and yappers).