I’ve been asked a lot how @MagicNewton ensures accuracy even during high market volatility. And it's a crucial thing to get.

To get this, you need to know how the internal process works.

It all starts with the user defining a rule like “Buy BTC if it drops below $100K.”

The AI agent then monitors on-chain data to check whether the condition is met. This monitoring occurs off-chain because it's resource-intensive and requires efficiency.

Now, you might say: if it's off-chain, can we trust it? The answer is in the next step.

Once the condition is met, the AI agent generates a ZK proof that it has followed the predefined rules and that the trigger condition is truly satisfied.

Since the first step is off-chain, the AI agent must prove on-chain that it didn’t cheat. That’s why it doesn’t matter that the AI operates off-chain, the trust comes from the on-chain proof.

A smart contract then verifies the ZK proof. If it’s valid and the original rule conditions are met, it authorizes the transaction. If not, it rejects it.

So, even in highly volatile conditions, nothing gets executed blindly. Right before the transaction goes through, the smart contract performs a final check.

If BTC has already jumped back above $100K, the contract's built-in revert logic cancels the transaction.

In short: the system is designed to remain accurate and trustless, even under extreme volatility.