Every week, someone loses money because they connected a wallet to a site, signed something they didn't understand and interacted with a contract that did exactly what its code said. You don't need to be a programmer to reduce that risk. You need a short routine of questions and an understanding of what a block explorer can and can't tell you.
This post is that routine, written for people who don't read Solidity. I'll point out where a developer would look at code, and what the same finding looks like in plain terms. At the end, I'll be honest about the limits: no checklist can promise that a contract is safe. It can only help you avoid the obvious traps. This is education, not advice.
Step 0: be sure you have the right address
Before analysing anything, confirm that you are looking at the contract the project really uses. Scammers deploy copies with the same name and logo. Get the address from the project's official website, its documentation or a link from a source you already trust, and compare the full address, not just the first and last characters. Never take an address from a message, a reply under a post or a search ad.
Step 1: is the source code verified?
Open the contract's page on the explorer and look at the Contract tab. A verified contract has its source code published, and the explorer has checked that compiling it produces the exact code deployed at that address. That is useful because it lets you, or someone you trust, read what the contract does.
It is also all that verification promises. A verified contract can still be badly written, unaudited, or designed to take your money. But an unverified contract is a page of hexadecimal that nobody can easily read. For anything where you'd put in real value, "unverified" is a reason to stop. If you're curious how verification works from the builder's side, I covered it in the Hardhat verification guide.
Step 2: is it a proxy, and who can change it?
Many contracts are proxies: a thin contract at the address you use that forwards every call to another contract, the implementation, where the real code lives. Etherscan explains the pattern in its help pages, including the "Read as Proxy" and "Write as Proxy" tabs that appear once the link is made. The reason projects use them is that the logic can be replaced without changing the address.
That is exactly the point to worry about as a user. If the logic can be swapped, then whoever holds the upgrade key can change what your funds are subject to. Two questions follow:
- Who can upgrade? A single wallet, a multisig of known people, or a contract with a delay (a timelock) that gives users time to exit?
- Is the implementation verified too? The explorer's own help page warns that it doesn't automatically verify that the code shown in the proxy tabs is what is actually being executed, so you should check the implementation yourself. The same page says that malicious implementations could be hidden from plain sight.
An upgradeable contract isn't bad on its own; many serious projects use them. But "can be changed by one person at any time" is a different risk from "cannot be changed". The UUPS vs Transparent post explains the mechanics for builders.
Step 3: what powers does the owner have?
Most contracts have an "owner" or "admin". On the explorer's Read tab, look for functions like owner, paused, isBlacklisted or cap. On the Code tab, look for the modifier onlyOwner and read what the functions behind it do. Here are the patterns that matter, with what each means for you:
- The owner can mint new tokens. More supply can dilute your holding. In the token from my ERC-20 tutorial, the owner can mint, but only up to a fixed cap enforced by the contract. A cap, a schedule, or no minting at all, are very different from unlimited minting.
- The owner can pause transfers or blacklist addresses. There are legitimate reasons (stopping an exploit), but it also means you may be unable to move your tokens.
- The owner can withdraw funds. In my simple NFT sale contract, the owner can withdraw the sale proceeds. That is expected for a sale, but a vault where the owner can take deposits is a different thing.
- The owner can change fees. Look for a fee setter with no upper bound. A fee that can be set to 100% is a trap waiting for a trigger.
- Ownership has been renounced. That removes owner powers, which can be reassuring, but also means nobody can fix a problem. And check it really happened: the
ownerfunction should return the zero address.
For each power, ask "who holds this, and could they abuse it?". If the owner is one anonymous wallet, assume the powers will be used.
Step 4: what are you actually being asked to sign?
Most losses don't come from the contract page; they come from the wallet pop-up. Before you confirm anything, read it:
- Is it a token approval? An approval lets a contract spend your tokens later. Many apps ask for an unlimited amount by default. If a site asks for one just to let you look at something, or to mint something, be suspicious. The token approvals post explains how to see and cancel them.
- Does the text match what you intended? If you wanted to swap and the pop-up says "set approval for all" on an NFT collection, stop.
- Is it a signature request rather than a transaction? Signatures cost no gas and can still authorize token movements in some designs. Treat unexpected signature requests with the same suspicion.
- Can you not read it at all? A wall of hex is a warning sign, not a formality. Ledger's documentation calls approving unreadable data "blind signing" and says it is where phishing and malicious contracts do their damage.
Step 5: look at the history, not just the code
- How old is the contract, and how many different users interact with it? A contract created yesterday with a handful of wallets is not the same as one that has run for years with many users.
- Has it been audited, by whom, and for which version? Look for a published report on the project's site and check that it covers the deployed version. An audit is not a guarantee, and a logo on a website is not an audit.
- Who is behind it? Anonymous teams aren't always fraudulent, but an anonymous team plus an upgradeable contract plus an owner who can withdraw is a high-trust combination.
- What do independent sources say? Search the project name with "exploit" or "scam" and read what comes up, from sources you can judge.
Step 6: reduce the damage if you're wrong
- Start small. Test with an amount you can afford to lose entirely.
- Use a separate wallet for experiments, holding only what you're willing to risk. Keep savings elsewhere.
- Approve exact amounts where the app allows it, and revoke afterwards.
- Never share a seed phrase or private key with anyone, ever. No legitimate site needs them to "verify" you.
What this checklist cannot do
- It can't find subtle bugs. A contract can be verified, owned by a multisig and still contain a logic flaw. Finding those takes expert review.
- It can't predict a rug pull by a trusted team. It only shows what powers exist.
- It can't make a risky idea safe. Unrealistic returns, pressure to act fast and promises of guaranteed profit are the warning signs, whatever the code says.
- Explorers can be wrong too. Etherscan's own page on proxies says that its proxy detection is heuristic and may be inaccurate.
The five-minute version
- Right address, from an official source.
- Verified source code, including the implementation behind any proxy.
- Who can upgrade, mint, pause, take funds or change fees?
- Read the wallet prompt: approval, signature or transaction, and for what amount?
- Start small, from a separate wallet, and revoke afterwards.
Building something people will trust?
I build contracts and dApps with clear permissions, tests and documentation, so users can check them. If you want a review of your own design, get in touch.
See services Get in touchWhat I ran this against
- Nothing in this post was tested by interacting with a live contract. The code examples are from the small Solidity contracts in this series (Hardhat 3.22.0, OpenZeppelin 5.6.1, run on 17 November 2026) and show what common risk patterns look like in source code.
- Block explorer features change. The descriptions come from Etherscan's own help pages and the sources below, read on 17 November 2026, and I did not navigate the live site for this post.
- This is education, not financial or security advice. I'm a developer describing how these tools work; I can't see your situation, and you remain responsible for what you sign.
Sources checked
- Etherscan Information Center, "Proxy Contracts" (how proxies delegate to an implementation, the "Read as Proxy" and "Write as Proxy" tabs, and its warning that Etherscan does not verify that the code shown is what is actually executed): info.etherscan.com.
- Etherscan Information Center, "Token Approvals" (reviewing and revoking what contracts may spend): info.etherscan.com/tokenapprovals.
- Zealynx Academy, "How to Read a Contract on Etherscan Without Being a Developer" (what "verified" does and doesn't mean; the Code, Read and Write tabs): academy.zealynx.io.