What exactly is @MagicNewton building?

🗂️Magic Newton is building a verifiable automation layer for onchain finance.

Their goal is to simplify the crypto landscape for users and unlock multiple opportunities through the use of AI agents

navigating the crypto space to profitability

Today the simplest of crypto tasks has to be done manually, and the more complex they are, the more dApps and chain we have to interact with

What if we can just ask a trustworthy agent to handle these tasks for us

- Say you want to buy a token at certain market conditions, how about an agent for that

- Or you want to find the best yield for your stables across multiple chains, how about an agent that can do that and manage your positions
MagicNewton would take the complexity out of crypto with not only chain abstraction, but also application abstraction

🔻Simplicity through Abstraction

With chain and application abstraction, users would not have to worry about what chain they're on, bridging etc

The Agents handles all the complexity for them, the user just has to select an agent and specify what they want

🔻Automation Marketplace

Bur more than just simplifying the crypto landscape, Newton is birthing a new economy, an automation economy

Let me explain.

If you go on http://newton.xyz, you'll see a "Recurring Buy Agent" which as the name implies can help you make recurring buys of a token

This is going to be one of the many agents that will be in Newton's "Automation Marketplace", developed by third-party devs as well as Magic

This serves to fuel an economy of opportunity for devs to build credible competitive agents.

But the question is, can we trust these agents?

🔻Are Newton's agents to be trusted?

Utilizing ZK-Tech, agents on Newton are bound by three principles

1. Scoped Autonomy: where users define the boundaries of agent behavior with zkPermissions
2. Verifiable Integrity: where agents generate cryptographic proofs to prove every action taken is as the user intended

3. Earned Reputation: where agents gain or lose reputation from their performance, alongside economic penalties to deter bad actors

With these in place, 
- users can trust agents to do tasks for them
- agents' performance is incentivized
- bad-actors are deterred
- an economy for devs is birthed, etc

The possibilities are endless, and good execution will be key