#newt $NEWT Peeling Off the Skin of AI Agents: What Is the Newton Protocol ($NEWT ) Up To?
Recently, the market has been going all-in on AI Agents. But after spending years wrestling with RPC endpoints and bare-metal servers, I only trust code logic: who would dare hand over control of funds to a black-box model? In this space, survival comes first.
I dug into the Newton Protocol’s GitHub repository and found its entry point is pretty crafty: it isn’t trying to compete for public-chain TPS. Instead, it hard-codes an “authorization layer” before transactions hit the chain. With cross-chain liquidity at the trillion-dollar scale, most of it is basically “unprotected”—direct execution via raw routing. Newton, in contrast, builds a Web3 version of Visa’s risk-control network.
Technically, it isn’t playing mysticism games. It uses TEE (Trusted Execution Environment) stacked with ZKP (Zero-Knowledge Proof). Compared with traditional DeFi that can only rely on static code audits for defense (e.g., cross-chain bridges that repeatedly fall victim to reentrancy attacks), Newton’s zkPermissions architecture forces AI agents to run inside a sandbox. It can intercept invalid actions at runtime—blocking execution immediately when risk-control conditions aren’t met. That turns “trusted institutions/code” into “validated cryptographic credentials.”
On fundamentals, the project is driven by Magic Labs, backed by more than $83 million in funding. Under the hood, it also conveniently plugs into EigenLayer for AVS node security verification, objectively stitching together the AI and Restaking narratives quite seamlessly.
However, the fatal flaw of infrastructure is always ecosystem inertia. For developers used to the EVM environment, will they be willing to pay the extra cost to adapt to its Rust-based Rego interpreter for writing strategies? With a total supply of 1,000,000,000 tokens—$NEWT units—and a token structure characterized by low circulation and high FDV, you still need safeguards to prevent a dump. As a breakthrough infrastructure for AI automation, it definitely deserves a place on the watchlist and close monitoring of how frequently its code gets submitted. But right now, calling it “disruptive” is just a harvest. Let it first run the mainnet loop end-to-end before we talk.
@NewtonProtocol
Recently, the market has been going all-in on AI Agents. But after spending years wrestling with RPC endpoints and bare-metal servers, I only trust code logic: who would dare hand over control of funds to a black-box model? In this space, survival comes first.
I dug into the Newton Protocol’s GitHub repository and found its entry point is pretty crafty: it isn’t trying to compete for public-chain TPS. Instead, it hard-codes an “authorization layer” before transactions hit the chain. With cross-chain liquidity at the trillion-dollar scale, most of it is basically “unprotected”—direct execution via raw routing. Newton, in contrast, builds a Web3 version of Visa’s risk-control network.
Technically, it isn’t playing mysticism games. It uses TEE (Trusted Execution Environment) stacked with ZKP (Zero-Knowledge Proof). Compared with traditional DeFi that can only rely on static code audits for defense (e.g., cross-chain bridges that repeatedly fall victim to reentrancy attacks), Newton’s zkPermissions architecture forces AI agents to run inside a sandbox. It can intercept invalid actions at runtime—blocking execution immediately when risk-control conditions aren’t met. That turns “trusted institutions/code” into “validated cryptographic credentials.”
On fundamentals, the project is driven by Magic Labs, backed by more than $83 million in funding. Under the hood, it also conveniently plugs into EigenLayer for AVS node security verification, objectively stitching together the AI and Restaking narratives quite seamlessly.
However, the fatal flaw of infrastructure is always ecosystem inertia. For developers used to the EVM environment, will they be willing to pay the extra cost to adapt to its Rust-based Rego interpreter for writing strategies? With a total supply of 1,000,000,000 tokens—$NEWT units—and a token structure characterized by low circulation and high FDV, you still need safeguards to prevent a dump. As a breakthrough infrastructure for AI automation, it definitely deserves a place on the watchlist and close monitoring of how frequently its code gets submitted. But right now, calling it “disruptive” is just a harvest. Let it first run the mainnet loop end-to-end before we talk.
@NewtonProtocol