Smart contracts: secure onchain agreements

A smart contract is a program stored on a blockchain that applies its predefined rules whenever a transaction calls it. Despite the name, smart contracts are not “contracts” in the legal sense, nor are they “smart” in terms of artificial intelligence. On Solana, they are called programs. Instead, they are computer scripts that replace a central intermediary, such as a bank, lawyer or escrow agent, with code. This allows complex financial transactions to execute transparently, without relying on a person to carry out the terms.

​

The mechanics: deterministic if/then logic

The fundamental power of a smart contract lies in its deterministic nature. In a traditional legal agreement, the outcome often depends on how the language is interpreted and on enforcement by a third party.

• A smart contract operates on strict logic: if event X occurs, then action Y is executed.
• Once deployed, it runs exactly as written: no one can make it skip a step or ignore its rules. Changing its behavior takes an upgrade, which only the holder of its upgrade authority can perform. On Solana, revoking that authority makes a program permanently unchangeable.

This logic allows for “programmable money”. For example, a smart contract can be programmed to hold a buyer’s funds in escrow and only release them to the seller once a shipping receipt can be independently verified. Because the terms are written into the blockchain, neither party can back out or delay payment once the conditions are fulfilled. The network itself acts as the impartial witness and executor of the trade.

​

Removing the middleman: trustless execution

In traditional finance, trust is expensive. When you trade stocks or buy property, you pay significant fees to clearing houses and brokers to make sure both sides fulfill their obligations. Smart contracts enable trustless execution: you do not need to trust the person you are trading with, only the underlying code and whoever can upgrade it. By acting as an automated escrow agent, a smart contract removes counterparty risk, the danger that the other person won’t deliver their end of the deal. When you swap on Jupiter, which routes your trade across AMMs and other liquidity sources on Solana, what you give and what you get move in the same transaction (atomic settlement). On Solana, if any part of a transaction fails, the entire transaction is reverted as if it never happened; only the network fee is still charged. Your capital is never left in a vulnerable, pending state.

​

Beyond simple transfers: use cases

While simple transfers are the baseline, smart contracts can be combined with each other to build entire decentralized ecosystems:

• Decentralized finance (DeFi): smart contracts act as automated market makers, lenders and insurers. Protocols can calculate interest rates from supply and demand, and let liquidators repay part of a borrower’s debt once a position crosses its liquidation threshold. The code enforces these rules, but people usually still set the parameters and run the liquidations: on Jupiter Lend, for example, interest rates adjust automatically, risk parameters are set per vault by an admin multisig behind a timelock, and liquidations are executed by third-party liquidators. See Liquidity Layer & Risk Management.
• DAOs (decentralized autonomous organizations): smart contracts can manage the governance of an entire organization. Votes are cast onchain, and if a proposal passes, the contract can move funds or change protocol parameters according to the result.

​

The trade-offs: “code is law” and its risks

A smart contract does exactly what its code says, which is its greatest strength and also its main vulnerability. This is often summarized as the “code is law” philosophy: any bug or exploit in the code is live as soon as the contract is deployed. If a developer leaves a backdoor or a logic error in the contract, an attacker can exploit it to drain funds, and there is no central reversal button to get the money back. Unlike a bank that can freeze a fraudulent transaction, a blockchain will faithfully execute the code exactly as it is written, even if the outcome is clearly unintended. Upgrades cut both ways. If a program cannot be upgraded, a bug in it can never be fixed. If it can, whoever holds the upgrade authority can fix bugs, but could also change the program’s behavior. Jupiter’s docs say which programs can be upgraded: the Jupiter Lock program is upgradeable, and each Jupiter Lend deployment (the Jupiter Market, and the Bitwise x Ethena / Sentora market) has its own upgrade multisig. This makes security audits, the process of having independent experts review the code before and after deployment, one of the most critical steps in the lifecycle of any onchain product. An audit lowers risk, but it does not guarantee that a program is free of bugs.

​

Go deeper