Transaction simulation is essential for wallet security, allowing users to see what will happen before they sign a transaction. However, existing methods rely heavily on third-party RPC providers to retrieve state data, creating significant trust issues. Users often find themselves in a blind signing situation, unable to verify what a transaction will actually do. Malicious actors can exploit this by providing false simulation results, leading to scenarios where users inadvertently authorize harmful transactions. To tackle this problem, the article proposes a trust-minimized simulation approach that utilizes Merkle Patricia Trie proofs combined with multi-node consensus. This method verifies the prestate before executing transactions, eliminating reliance on any single RPC provider. The process involves fetching prestate data from multiple independent nodes, requiring unanimous agreement on the state root, and verifying all necessary proofs locally. Only after confirming these proofs does the system execute the transaction using a local EVM implementation, generating its own execution traces. The approach addresses the vulnerabilities of compromised RPC providers and simulation services, which can be targeted through various attacks like DNS hijacking or infrastructure breaches. By decentralizing the verification process and ensuring that all nodes must agree on the data, the system significantly reduces the risk of fraudulent simulations. If any discrepancies arise during the verification, the simulation is rejected outright, maintaining a high level of security for users’ transactions.