It's so hard to triage this when no justification has been provided for the advisory. Was the GHSA released in response to npm pulling the package, or vice versa?
Many suggestions for workarounds, but if the GHSA is indeed accurate (all versions affected) then that seems unwise.
Also if all the versions are affected this malware is in stylus since 2010. Honestly, it sounds improbable to me that a malware exists unnoticed in open source software for 15 years. However, even if improbable it's better to play safe and just override the installation of stylus ( especially if you are not using it ) with an empty package until more information is released
My staging build was failing and I saw that stylus was the culprit. Running `npm why stylus`, `npm ls --all stylus`, and other variants of these two commands consistently returned nothing, but I can see it in my lockfile if I run `grep -R stylus package-lock.json`.
Even running `npm audit | grep stylus` returned nothing! Which I think is pretty crazy considering the package itself has been overwritten by NPM to include a 0 context scary "Security holding package" thing. Surely this sort of thing should show up in the `audit` results?
I have to say NPM packaging is terrible. I probably spend 1 month of the year fiddling with upgrading packages due to security issues. That is just the amount of time I spend on my repos alone. All of this extra effort to avoid code signing and making package owners accountable.
It seems like every week there is a new security high sev ticket to fix some webpack dependency.
Not to mention that even if you do successfully run “npm audit fix” (—force), Npm may not update to the correct new version and will often downgrade packages many many many versions.
The error messages that Npm spits out have always frightened junior devs too.
I can’t wait for that whole ecosystem to be replaced.
I see two comments here on this subject, complaining about the churn of dealing with security advisories. Sure, it's churn.
... but isn't this problem dwarfed by the implications of having used a compromised package? Presumably, if the project you work on has a compromised dependency, it means you've ran it on your development machine. Presumably, you might have a couple of secrets (private keys, AWS credentials and other whatnots) lying around, which might have leaked to a malicious actor.
Wouldn't you need to review all the development, staging and production machines for all your projects and rotate secrets everywhere?
Wouldn't it be, by far, the biggest churn involved, so much that mentioning "npm audit" difficulties not worth mentioning at all, because of the ridiculous comparison in effort magnitude?
I think this would be a fair assessment, if the security advisory would be true.
Since it's most probably false, the implications you refer to remain hypothetical, while the cost of cleaning up after npm's decision are measured in real M$s. And I think that's the real issue here.
I am not saying that we should give up on security altogether, but now there is so much toil attached to managing security, compliance and such aspects of the development lifecycle, that at some point managing all these aspects will outweigh all productivity a dev can bring to the project.
It's admittedly a hyperbole, but at that point the whole development procedure would simply become a pointless exercise without any benefit to anyone.
In my case, stylus is a transient dependency of a transient dependency of a transient dependency... Vite has had stylus as an OPTIONAL peer dependency for a very long time now, and stylus itself has existed for MANY years.
What NPM did here is eradicate every single version of stylus ever published, so the breakage for the large majority of people here is that NPM will now try to fetch a non-existent package, which will cause CI and other scripts that rely on `npm ci` or `npm i` to fail.
It's one thing to get a big scary warning saying "Hey, stylus has a vulnerability, here's an overview of the issue..." and then pushing out the overwritten version as its own standalone version that people can migrate to. Instead, NPM silently overtook a package and overrode it completely. Running `npm audit` in a project affected by this, I see 0 mention of stylus in it, there is ZERO indication anywhere that something about this package is wrong other than the fact that the package basically doesn't exist in the registry anymore. And in my testing so far, things like `package.json` `overrides` fields does not work [1].
So I wouldn't say this is your typical vulnerability situation. They pulled packages with 0 warning or notice to anyone, and their own security audit tooling gives you nothing to go by, and there seems to be basically nothing you can do to fix this, depending on how exactly your project is setup. We're not even sure there is an actual attack or vulnerability, because they don't link to any details literally anywhere! Just take a look at the NPM page [2], there are ZERO details here! And even weirder (could be that NPM just doesn't count downloads this early into a change's lifecycle), the downloads for the version they override is sitting at 0, to me indicating that nobody has been able to even download this, which I can confirm at least anecdotally from me trying to fix this issue myself.
No the biggest churn involved is now I’m another engineer that prefers to stay away from using, developing on, and recommending javascript platforms.
To your point I think you will find most companies stop at the upgrade high sev packages step and do not have any requirements or churn related to checking for fallout from sevs.
The crazy thing is that `npm audit` doesn't even list `stylus` here, at least not in my repos. Despite them literally overtaking the damn package on the registry for a *security issue*.
A quick workaround if you're affected by a deep dependency and don't rely on stylus directly - add `"overrides": {"stylus": "0.0.1-security"}` to your package.json
From how is unfolding the most probable outcome is that one of the maintainer is compromised ( Ponya ), all of the packages he contributed to have been marked
Amateur hour all around in that thread.
I can't believe that people are actually, unironically recommending that you use a mutable git tag reference in package.json when they should be using a tamper-proof git SHA instead.
Work around the issue by installing directly from GitHub using package.json overrides:
```
"overrides": {
"stylus": "github:stylus/stylus#0.64.0"
}
```
The title is wrong. There's no proof of compromise. There are no releases of the package since October. Apparently one of the long-time maintainers has pushed other compromised packages, so npm just nuked all the packages he had access to, whether they were compromised or not.
Readit News
https://github.com/advisories/GHSA-fh4q-jc76-r59p
I'm still unsure if it's a mistake on NPM side or if stylus and the authors are compromised
Many suggestions for workarounds, but if the GHSA is indeed accurate (all versions affected) then that seems unwise.
And the GHSA advisory: 2025-07-23T03:03:56Z
So the GHSA was released after the pull (by a minute).
My staging build was failing and I saw that stylus was the culprit. Running `npm why stylus`, `npm ls --all stylus`, and other variants of these two commands consistently returned nothing, but I can see it in my lockfile if I run `grep -R stylus package-lock.json`.
Even running `npm audit | grep stylus` returned nothing! Which I think is pretty crazy considering the package itself has been overwritten by NPM to include a 0 context scary "Security holding package" thing. Surely this sort of thing should show up in the `audit` results?
Dead Comment
It seems like every week there is a new security high sev ticket to fix some webpack dependency.
Not to mention that even if you do successfully run “npm audit fix” (—force), Npm may not update to the correct new version and will often downgrade packages many many many versions.
The error messages that Npm spits out have always frightened junior devs too.
I can’t wait for that whole ecosystem to be replaced.
I see two comments here on this subject, complaining about the churn of dealing with security advisories. Sure, it's churn.
... but isn't this problem dwarfed by the implications of having used a compromised package? Presumably, if the project you work on has a compromised dependency, it means you've ran it on your development machine. Presumably, you might have a couple of secrets (private keys, AWS credentials and other whatnots) lying around, which might have leaked to a malicious actor.
Wouldn't you need to review all the development, staging and production machines for all your projects and rotate secrets everywhere?
Wouldn't it be, by far, the biggest churn involved, so much that mentioning "npm audit" difficulties not worth mentioning at all, because of the ridiculous comparison in effort magnitude?
Since it's most probably false, the implications you refer to remain hypothetical, while the cost of cleaning up after npm's decision are measured in real M$s. And I think that's the real issue here.
I am not saying that we should give up on security altogether, but now there is so much toil attached to managing security, compliance and such aspects of the development lifecycle, that at some point managing all these aspects will outweigh all productivity a dev can bring to the project.
It's admittedly a hyperbole, but at that point the whole development procedure would simply become a pointless exercise without any benefit to anyone.
What NPM did here is eradicate every single version of stylus ever published, so the breakage for the large majority of people here is that NPM will now try to fetch a non-existent package, which will cause CI and other scripts that rely on `npm ci` or `npm i` to fail.
It's one thing to get a big scary warning saying "Hey, stylus has a vulnerability, here's an overview of the issue..." and then pushing out the overwritten version as its own standalone version that people can migrate to. Instead, NPM silently overtook a package and overrode it completely. Running `npm audit` in a project affected by this, I see 0 mention of stylus in it, there is ZERO indication anywhere that something about this package is wrong other than the fact that the package basically doesn't exist in the registry anymore. And in my testing so far, things like `package.json` `overrides` fields does not work [1].
So I wouldn't say this is your typical vulnerability situation. They pulled packages with 0 warning or notice to anyone, and their own security audit tooling gives you nothing to go by, and there seems to be basically nothing you can do to fix this, depending on how exactly your project is setup. We're not even sure there is an actual attack or vulnerability, because they don't link to any details literally anywhere! Just take a look at the NPM page [2], there are ZERO details here! And even weirder (could be that NPM just doesn't count downloads this early into a change's lifecycle), the downloads for the version they override is sitting at 0, to me indicating that nobody has been able to even download this, which I can confirm at least anecdotally from me trying to fix this issue myself.
[1] https://github.com/npm/cli/issues/4232 [2] https://www.npmjs.com/package/stylus
The vast majority of "compromised packages" are just dev dependencies that have a slow regexp.
To your point I think you will find most companies stop at the upgrade high sev packages step and do not have any requirements or churn related to checking for fallout from sevs.
So you probably need to carefully audit the changes from two data sources and the security ticket ends up being 2+ merge requests.
Maintainer @iChenLei reports they are negotiating with npm officials to restore access: https://github.com/stylus/stylus/issues/2938
This has been reflected in a recent edit and comments here: https://github.com/stylus/stylus/issues/2938
No updates to the security advisory at this time: https://web.archive.org/web/20250723155624/https://github.co...