Malicious Browser Extensions And Clipboard Hijackers On A Farming Machine
Losses with no bad signature in them
Most wallet security advice is about the moment you sign. Read the transaction, simulate it, check the contract, don’t approve unlimited spending, don’t sign messages you can’t decode. That advice is right, and it assumes the machine in front of you is telling the truth.
This piece is about what happens when it isn’t. A malicious browser extension can change what a dapp page shows you before you ever see the wallet popup. A clipboard hijacker can replace the address you copied with a different one in the fraction of a second between copy and paste. An information stealer can lift your encrypted wallet vault and saved passwords straight off the disk. In each case, you might do everything “right” at the signing step and still lose funds, because the thing you were checking was already compromised.
The honest framing is this: a clean signature on a dirty machine is still a risk. The signing habits covered elsewhere on this site protect you against malicious sites and contracts. They do much less against malware sitting inside your own browser or operating system.
Farming machines are unusually exposed. They tend to run many browser profiles, often with helper extensions installed for convenience. They’re used to download tools, scripts and bots from places with loose standards. They sit in dozens of Discord and Telegram communities where malicious links circulate. And they usually hold many wallets at once, which makes one infection far more expensive than it would be on a normal laptop.
What follows covers the main ways endpoint malware takes funds, how to audit what’s installed, how to verify you’ve got the real extension, and why keeping a deliberately lean signing profile does more than any single tool. It extends the separate-machine piece on this site, which covers where signing should happen, with the question of what else is running there.
Why extensions are so powerful
A browser extension is a small program the browser runs with permissions you granted when you installed it. Those permissions can be very broad, and the install prompt describes them in language most people skim.
The permission that matters most is the one usually described as being able to read and change all your data on all websites. With it, an extension can inject scripts into every page you open. On a dapp page, that means it can read what’s shown, change what’s shown and interact with the page as if it were part of it. It can alter the destination address in a send form after you’ve pasted it. It can change the amount. It can add a hidden element that triggers a transaction request which looks like it came from the site.
Extensions can also request clipboard access, which lets them read what you copy and write something else in its place. They can request access to your browsing history, your tabs and your downloads. Some can talk to native programs on your machine.
What an extension generally can’t do is reach directly into another extension’s private storage, which is where your wallet extension keeps its encrypted vault. Browsers isolate extensions from each other. So a malicious extension usually works around the wallet rather than breaking into it: by manipulating the pages you interact with, by stealing what you type or copy, or by tricking you into approving something the wallet faithfully shows you.
That’s worth keeping in mind, because it means the wallet popup is still one of your best checks. If the popup shows a destination or an action you didn’t expect, believe the popup over the page. The page is where a malicious extension does its work.
Copycat wallet extensions
The most direct attack is to get you to install a fake wallet in the first place.
Fake wallet extensions imitate the name, logo, screenshots and description of popular wallets. They appear in official extension stores, sometimes for weeks before they’re removed, and sometimes with inflated ratings and reviews. They appear in search ads, where a paid result above the real wallet’s site leads to a lookalike download page. And they get linked from forums, videos and chat groups as “the updated version” or “the version that supports the new chain”.
A fake wallet doesn’t need to be subtle. When you import your seed phrase, it sends it to the attacker. When you create a new wallet, it may generate a seed the attacker already knows. Either way, everything in that wallet is theirs from the start, and the theft may be delayed so the connection to the install isn’t obvious.
The defence is to never find a wallet extension by searching the store or clicking an ad. Go to the wallet’s official website, reached from a source you trust, and use the store link it provides. Then check that what you’re installing is the listing that link points to. Store listings have unique identifiers, and many wallets publish their extension’s ID on their site. If the listing you’re looking at has a different ID from the one the official site gives, it’s not the real extension, whatever the name says.
Look at the publisher and the listing history too. A real wallet extension has usually been in the store for a long time, with a large user base and a publisher name that matches the company. A fake one might have a publisher name that’s close but not identical, a short history and reviews that read like they were written together.
When the real extension goes bad
A harder problem is when a legitimate extension becomes malicious. This happens in a few ways, and none of them require you to do anything wrong.
The first is a compromised developer account. Extension stores push updates automatically. If an attacker gets into the account that publishes an extension, whether through phishing or stolen credentials, they can publish a malicious update, and every user with that extension installed receives it without being asked. At the end of 2024, a batch of legitimate Chrome extensions, including one from a security company, pushed malicious updates after their developers were phished through a fake store notice. The affected extensions weren’t wallet extensions, but the mechanism works on any extension.
The second is a sale. Popular extensions with many users are valuable, and developers sometimes sell them. The new owner inherits the users and the update channel. Some buyers are legitimate. Others buy extensions specifically to push adware, tracking or worse to an existing user base.
The third is a compromised dependency. An extension may include third-party code that gets updated upstream, and if that upstream package is compromised, the extension can carry the problem even if its own developer did nothing wrong.
This is why “only install trusted extensions” is incomplete advice. An extension that was trustworthy last month can change. The better goal is fewer extensions, in the places that matter. An extension you don’t have can’t be updated into something harmful.
Some browsers let you see when an extension last updated and what permissions it holds. If an extension suddenly asks for new permissions after an update, the browser will usually disable it until you approve. Treat that prompt seriously. A colour picker that now wants access to all websites is a reason to remove it.
How clipboard hijackers work
Clipboard hijackers, sometimes called clippers, are one of the oldest tricks in crypto malware, and they’re still effective because they exploit a habit everyone has.
The malware runs in the background on your computer. It watches the clipboard. Whenever something that looks like a crypto address gets copied, it replaces it with an address the attacker controls. You copy your exchange deposit address, switch to your wallet, paste, and the field now holds the attacker’s address. If you don’t compare the pasted address with the source, you send funds to them.
Addresses have recognisable formats, which makes this easy to automate. An Ethereum-style address has a fixed length and character set. Bitcoin, Solana and other chains have their own patterns. A clipper can recognise all of them and keep a replacement ready for each.
Some clippers go a step further and pick a replacement that looks like the address you copied, matching the first and last few characters. That’s the same trick used in address poisoning, and it means a quick glance at the ends of the pasted address may not catch it.
Clippers can also live in browser extensions with clipboard permission, though the classic versions are operating-system malware. Either way, the effect is the same: the thing you paste isn’t the thing you copied.
Nothing about this involves a malicious signature. Your wallet shows you a normal transfer to an address, and you approve it. The wallet is working correctly. The address was wrong before it ever reached the wallet.
Information stealers and the wallet vault
Clippers wait for you to make a transfer. Information stealers don’t wait.
A stealer is malware that collects valuable data from a machine and sends it to the attacker, usually within minutes of infection. Crypto-focused stealers typically look for browser profile folders, where wallet extensions keep their encrypted vaults, and for saved browser passwords, cookies and session tokens. Some also look for desktop wallet files, text files with names like seed or wallet, screenshots and password manager data.
A wallet extension’s vault is encrypted with your wallet password. That’s real protection, but it’s only as strong as the password. If the password is weak, reused or saved in the browser, a stolen vault can be decrypted. Attackers try stolen passwords from the same machine first, which is why saving your wallet password in the browser’s password manager undoes most of the vault’s protection.
Stolen session cookies are a separate risk. With them, an attacker can sometimes access your exchange account, email or chat accounts without your password, because the session is already logged in. That can lead to changed withdrawal addresses, taken-over community accounts and further attacks on people who trust you.
Stealers arrive the usual ways: cracked software, fake installers, malicious attachments, and increasingly the “verification” trick where a fake captcha tells you to paste a command into your operating system’s run dialog. For farmers, a common vector is tools: an “airdrop bot”, a “multi-wallet manager” or a “gas optimiser” shared in a chat group or on a code site, promising automation and delivering a stealer.
If a stealer has run on a machine, assume anything it could read is gone. That means moving funds with new seeds generated somewhere clean, not just changing passwords.
Why farming machines attract this
Anyone with crypto on a computer can be infected. A few habits common to farmers raise the odds and the stakes.
Farmers download tools. Automation scripts, trackers, bots and helper programs are part of the work for many people, and plenty of them come from small developers, anonymous repositories or chat groups. Each download is a chance to run someone else’s code on the machine that holds your wallets.
Farmers install helper extensions. Quest trackers, gas estimators, price overlays, tab managers, auto-clickers, translation tools. Each one is useful on its own. Each one is also code with broad permissions and an update channel you don’t control.
Farmers run many profiles. An anti-detect browser or a set of browser profiles lets you keep wallets separate, but it also multiplies wherever extensions get installed. Some tools let you install an extension across many profiles at once, which is convenient until that extension receives a bad update and it reaches every profile at the same time.
Farmers sit in many communities. Every Discord and Telegram group is a source of links, files and “verification” bots, some of them malicious.
And farmers hold many wallets. On a normal laptop, one infection might cost one wallet. On a farming machine, one stealer can collect every vault in every profile, and one clipper can intercept every paste across every wallet.
None of this means farming is reckless. It means the endpoint needs the same deliberate structure as the wallets themselves.
One helper extension across twenty profiles
A made-up but realistic example shows how the scale problem plays out.
Someone runs twenty wallets, each in its own browser profile inside an anti-detect tool. Months ago, they found a handy extension that shows quest progress and gas prices in a small overlay on dapp pages. It saved them a few clicks per task, so they installed it in all twenty profiles using the tool’s bulk install feature.
The extension’s developer, a single person maintaining it in their spare time, gets an email that looks like a store policy warning. It links to a page asking them to re-authorise their developer account. They do. A few days later, a new version of the extension is published from their account. Nobody was asked. Every profile updates automatically overnight.
The new version adds a small script to dapp pages. When a send or swap form appears, it quietly swaps the destination field to an attacker’s address after the user pastes. It doesn’t touch every transaction. It waits for larger amounts.
The farmer notices nothing for a week. Then, after a claim, they sweep proceeds from several wallets to their holding address. Some of those sweeps go to the attacker instead. The wallet popups showed the attacker’s address each time, but the farmer, moving quickly through a batch, checked the dapp page rather than the popup.
What would have stopped it? Several things. Not having the extension in signing profiles at all. Restricting its site access so it couldn’t run on dapp pages. Reading the destination in the wallet popup or on a hardware wallet screen. Picking the holding address from an address book instead of pasting. Any one of them would have been enough.
What wouldn’t have stopped it: installing only from the official store, since this was the real extension; checking the publisher, since the publisher hadn’t changed; or antivirus, since the malicious code lived in a browser extension update.
Profile separation versus machine separation
It’s worth being clear about what browser profiles do and don’t isolate, because it’s easy to assume more separation than you have.
Browser profiles, including anti-detect profiles, separate cookies, local storage, extensions and fingerprint details from one another. An extension installed in one profile doesn’t run in another. That’s real isolation for browser-level threats, and it’s why a lean signing profile is worth having.
Profiles don’t separate the operating system. A clipboard hijacker running on the machine sees every paste from every profile. An information stealer can read the files of every profile on the disk, including the wallet vaults inside each one. A keylogger captures every password typed in any profile.
So there are two layers of separation, and they defend against different things. Profile separation limits the damage from a bad extension. Machine separation limits the damage from operating-system malware. If you only have profiles, a single infection at the system level reaches everything.
For a small farm, the reasonable middle ground is a signing machine that runs nothing but the browser and wallets, with research and tools on a different device. The separate-machine piece on this site covers the setup. For larger operations, the case for splitting high-value wallets across separate physical machines gets stronger as the total held grows.
Auditing the extensions you already have
Start with what’s installed today. This takes less time than people expect.
Open the browser’s extension manager in every profile you use, including profiles inside anti-detect tools. For each extension, ask four questions. Do I still use it? Where did I get it? Who publishes it? What can it access?
Remove anything you don’t actively use. This is the highest-value step. Most people find extensions they installed once for a single task and forgot about.
For the ones you keep, check the publisher and the store listing. Is it the same publisher you think it is? Has it changed hands? Does the listing’s history look normal? If an extension was installed from a file rather than the store, sometimes shown as “unpacked” or installed in developer mode, and you don’t remember doing that, treat it as suspicious.
Check permissions. Many browsers let you restrict an extension’s site access to “on click” or to specific sites, instead of all sites. For extensions that don’t need to run everywhere, restrict them. An extension that only needs one site shouldn’t be able to read your dapp pages.
Check incognito and private mode settings. Extensions that are allowed in private windows run there too.
Finally, check the wallet extension itself. Confirm its store ID matches the one published on the wallet’s official site. It’s rare for a legitimate install to be replaced, but checking once removes the doubt.
Put a reminder in your calendar to do this again every month or two. Extensions change. Your list should be short enough that checking it takes five minutes.
Keeping a lean signing profile
The single most effective structural change is to separate where you sign from where you do everything else.
A lean signing profile is a browser profile that contains the wallet extension and nothing else. No trackers, no overlays, no productivity tools, no translation helpers. You use it to open dapps and sign. You don’t browse random sites in it, you don’t click links from chat groups in it, and you don’t install “just one more” helper in it.
The logic is simple. Every extension in the signing profile has a chance to interfere with the pages you sign on. If there are no other extensions, there’s nothing to interfere. The browser still has its own risks, but the extension surface drops to the one piece of software you actually need.
For multi-wallet setups, this scales naturally. Each wallet’s profile carries its wallet extension and whatever isolation the profile provides, and nothing else. Research, community browsing and tool usage happen in separate profiles or on a separate machine.
Some people find this inconvenient because they like having a quest tracker next to the dapp. The trade-off is real. A tracker in a separate profile, or on a second screen, costs a few seconds per task. An extension compromise in your signing profile costs whatever’s in the wallets there.
If you use an anti-detect browser to manage many profiles, check how it handles extensions. Some install extensions per profile, some share them, and some let you push to many profiles at once. Know which behaviour you have, and keep the signing profiles deliberately minimal.
Defeating the clipboard swap
You can’t always know whether a clipper is running. You can make its swap fail.
Compare every pasted address with its source, checking a chunk from the middle as well as the ends. A clipper that picks a lookalike matches the ends. It doesn’t usually match the middle.
Use your hardware wallet’s screen. When a hardware wallet signs a transfer, it shows the destination address on its own display. That display is driven by the device, not by your computer, so malware on the computer can’t change what it shows. Reading the destination there, including a middle chunk, catches a swapped address at the last possible moment. For any meaningful amount, this is the most reliable check available to you.
Use address books and whitelists. If you pick a saved destination by name in your wallet, instead of pasting, there’s nothing for a clipper to replace. Exchange withdrawal whitelists do the same on the exchange side. They can’t protect against malware that changes the saved entry itself, but they eliminate the paste step for routine transfers.
Be suspicious of any mismatch. If the address in the wallet doesn’t match what you copied, stop. Don’t assume you copied the wrong thing. Check the machine.
And be aware that clippers aren’t limited to addresses. Some replace other copied data too. If you ever copy a seed phrase, which you shouldn’t need to do, a stealer or clipper can capture it.
Verifying downloads and tools
Tools are the main way malware reaches farming machines, so they deserve a process of their own.
Prefer tools with a real track record: a named developer or company, a public repository with history, documentation and an active issue tracker. A tool that exists only as a file in a Telegram group, or a repository created last week with a single commit and a download link, is a much higher risk.
If a tool is open source, look at what it does before running it, or at least at what it asks for. A script that asks for your private keys in a config file, needs administrator rights, or downloads further code from somewhere else at runtime needs a very good reason to exist.
Never run tools on your signing machine or in your signing environment. If you need to test something, use a separate machine, a virtual machine or a disposable environment that holds no wallets and no logged-in sessions.
Watch out for “cracked” or “free premium” versions of paid tools. They’re one of the most common ways stealers get distributed, precisely because the people downloading them are motivated to ignore warnings.
Be especially careful with anything that asks you to paste a command into a terminal or run dialog. That’s the entire delivery method for a whole family of stealers. Legitimate installers don’t come as a line of text a stranger tells you to paste.
What to do if you suspect an infection
If you notice a pasted address that doesn’t match, an extension you don’t recognise, or funds moving without your action, act on the assumption that the machine is compromised.
Stop using it for anything wallet-related immediately. Don’t unlock wallets on it. Don’t enter passwords on it.
From a different, clean device, create new wallets with new seed phrases. This is critical: a new seed generated on an infected machine may be captured at the moment it’s created. Then move funds from the exposed wallets to the new ones. If the old wallets’ seeds were on the infected machine in any form, including inside a wallet extension’s vault, treat them as known to the attacker and move quickly, starting with the largest balances.
Revoke approvals on the old wallets where you can, since an attacker who knows a seed can use any approvals those wallets hold.
Change passwords and log out all sessions for exchanges, email and chat accounts that were accessed from the machine, again from a clean device. Enable or re-check two-factor authentication.
Then deal with the machine itself. For most people, the safest course after a real infection is a full wipe and reinstall, since stealers and clippers can hide persistence mechanisms that are hard to find.
Finally, work out how it happened, if you can. The tool, the extension, the link, the paste-this command. The point is the next machine.
What alternatives get wrong
Some common approaches sound protective but don’t address the problem.
“My antivirus would catch it.” Antivirus catches known malware. Crypto stealers and clippers change frequently, and a new variant may not be detected for a while. Antivirus is worth having. It isn’t a substitute for not running untrusted code.
“I only use extensions from the official store.” Stores remove bad extensions, but fakes can sit there for a while, and legitimate ones can be compromised through updates. The store tells you where the extension came from. It doesn’t guarantee what it does next month.
“My wallet has a password, so a stolen vault is fine.” That depends on the password. If it’s saved in the browser, reused or weak, it probably isn’t fine.
“A hardware wallet makes all of this irrelevant.” A hardware wallet protects keys from being stolen and gives you a trusted screen. It doesn’t stop you approving a transaction to a swapped address if you don’t read the screen, and it doesn’t protect browser sessions or exchange accounts.
“I’ll use a burner machine for everything risky.” That’s a good idea in principle. It works only if the burner really holds nothing valuable, including no logged-in sessions and no password manager.
Where this advice has limits
I want to be honest about what these steps do and don’t achieve. They reduce the chance of an infection reaching your signing environment and make the most common attacks fail at the point of sending. They don’t make a machine immune.
Browsers themselves occasionally have serious vulnerabilities, and a malicious site exploiting one can bypass the extension question entirely. Keeping the browser updated matters more than any extension audit.
Operating systems have the same issue. A fully patched system with minimal software is harder to compromise than one running old versions with lots installed.
And the human part never goes away. The most careful setup can be undone by one paste-this command in a moment of impatience, or one helper extension installed in the signing profile “just this once”.
I haven’t measured infection rates on farming machines and I don’t know of reliable public figures for that specifically. What’s clear from the mechanics is that every extension, tool and download adds surface, and the machine that signs your transactions should carry as little of it as you can manage.
For more on wallet security, separate signing setups and the tools we actually trust on a farming machine, head back to the Airdrop Farming homepage.
Get new guides and videos first — join the Telegram channel.