Hacker Newsnew | past | comments | ask | show | jobs | submit | damienwebdev's commentslogin

After yesterday's composer situation, I felt the need to write down my perspective on how this came to be as the vulnerability reporter.


It's a mix of responsibility between both systems. A new format combined with an old "bug".


I was the reporter on this one. If you have Github Actions in your organization, disable them immediately if you're unsure which version of composer your Github Actions run.


Could you elaborate on this bit on why Github's secret masking doesn't work here:

> GitHub Actions' built-in secret masker matches registered values as exact substrings. When the exception message is rendered by Symfony Console it may wrap, embed in In BaseIO.php line N: framing, or interleave with ANSI control sequences. So the masker does not redact, and the plaintext token reaches the log.

What does this log rendering look like such that the token from that code snippet becomes interleaved enough to not be a substring match?

I'm not familiar with composer/Symfony but I would expect something like:

  Line 34: Foo bar
  Line 34: <red>Foo bar</red>


From what I can tell Symfony wraps these errors in a styled error block so it messes with GitHub's masking. For example, newlines breaking up the token are apparently enough to trick the masker since it's not a substring anymore.

For your example:

Line 34: <red>ghs_fdsafdsarewqfdsa...</red>

Is probably long enough in most cases to break to a new line.

Also, those blocks get "interleaved" like this:

<red>ghs_fdsafdsarewqfdsafdsa</red>

<red>fdsardsardsafdsafs......</red>

So there are now </red>'s within the token, also breaking the substring matching.

This is kind of eye-opening for me -- I have never thought about how styled output like this could cause logging leakages.


Aah, the newlines were the thing I was missing. That makes sense then.


Nice find. The tokens being leaked in actions log was not one of the security implications I thought of when they released the feature.

How many other actions/libraries do you think are vulnerable?


I don't have an exact count, but during my analysis, I found that ~40 of the top 100 starred repos in the PHP ecosystem were impacted. Primarily by jobs that run `on: schedule` or by a maintainer with an `on: push`


Not immediately clear to me, is this limited to ghu_xxx type OAUTH tokens? And it's only relevant for PHP projects that use composer in GHA?


That's my understanding. This seems to only effect PHP projects that use Composer in GitHub actions. Examples being usage of shivammathur/setup-php and or php-actions/composer.


It's limited to ghs_ (server to server token's), that have the new format enabled: https://github.blog/changelog/2026-04-24-notice-about-upcomi... (and actions that use the vulnerable package)

This include's the GITHUB_TOKEN that is builtin within a actions jobs.


From a brief review, it looks like the underlying platform they use is https://www.scayle.com/ (though I'm not sure its the one that was attacked) its just the one I found while looking at their site.


I'd like to think that I'm a somewhat "odd" person, but I'm not exactly sure how that reflects in my code. This code hasn't been taken over, it's just been something I've been working on a long time.


I noticed this too. I was actually working on this yesterday, https://github.com/graycoreio/daffodil/pull/3946

I got a little side-tracked over the weekend and didn't get a new build of the demo in with the new image loader. I'll have this fixed shortly!


Just following up that I've fixed this!


That was fast! Works a lot better on my browser now ;)


There are many tech startup companies that pay for:

1. HN Upvotes 2. Product Hunt Upvotes 3. Github Stars

Why? Because it's all part of their marketing con. For some software companies, they rely on the hivemind to drive the purchase of their vaporware.


I'm one of the maintainers of Faker, so hopefully I can help remedy any situation we inadvertently created.

https://fakerjs.dev/api/helpers.html#arrayelement

I believe is what you're looking for. The issue, as I recall, is that all of Faker is random, so what does it mean to have a `random` module namespace. We moved it out of a hope for improved clarity.


Oh shit, youre right. Thanks for that. I think I saw that method but didnt realize it is implicitly random (like an idiot)


So, I happen to know this niche extremely well!

I'm one of the maintainers of MageOS (a fork of Magento) as well as a maintainer of Daffodil (a monolithic Angular PWA framework - not Microfrontends yet).

APIS are definitely the way to go but, while that's a great next step, what we're really facing is a standards issues.

Some have tried.

Schema.org as a standard is overly complex and each platform has their own API definition.

Microfrontends can be great as well, but ecommerce has a particularly special problem of requiring extremely performant SSR, so that's always a critical and complex piece to get correct.

However, this article is clearly an ad.


Former M1/M2 Dev here. The benefits of headless to me are that it's far easier to find javascript developers than to ruin someone with learning the Magento architecture.

I haven't seen Magento - under Ebay, under Adobe, under anyone address what I think is the biggest issue facing them - total cost of ownership in comparison to the competition.

Is mageOS mostly just a maintainence mode, keep this alive till everyone can switch off (like the M1 forks that exist) or do you actually envision trying to solve some of the challenges magento is facing as a community and platform?


MageOS is not a maintenance mode, unfortunately we don't do a good job expressing that.

The goal is to migrate management of the codebase out of Adobe's hands back into the community's. Community PRs, triage, etc.

We're currently working on getting the entire CI pipeline setup in Github Actions (if you're interested in contributing, we could really use friendly faces, it's quite an undertaking). Once we're past that, we have goals that are in progress:

DevX:

1. Correct the dep graph of MageOS to be a legitimate DAG and then use nx/bazel/make, etc to compute the build graph. 2. Create codespaces/gitpods/devcontainers for easier day 1 devs. 3. Rip GraphQl out into a separate package space. 4. Create smaller more use-case specific metapackages so that you're not forced into all of Magento or nothing. 5. Maintain backwards compatibility with existing Magento packages.

UX: 1. Significantly improve overall performance.

MerchX: 1. Improve Merch documentation and make Merch docs easier to consume.

Generally to your question, we're not interested in maintenance, we're interested in taking on the burden of being a better development effort alongside a better merchant experience.

That said, this is not a sales pitch. The problem is hard, but we'd like to prove that we're harder.


The problem space is huge. It's akin to LibreOffice, we have this huge amount of code, written by enterprises in complex ways, and the end user doesn't care, they just want it to work.

The backwards compatibility is really a challenge, but I think more importantly dumping Knockout and Luma and all that insanity and going towards the Hyva route, or going more-headlessy with a good starter theme would encourage people enourmously.

Magento's frontend cluster-F is the reason we moved away from it. It was great to have such a good pricing engine, integrations pieces, currency support etc, but all of that skillset which is magento's core capability being a really well rounded online OMS is overshadowed by long and expensive turn arounds on frontend work.


Fully agree. The problems are hard, the interested talent is limited (who would want to work on line of business software when academia actively pushes people towards other things), and the problem space (all possible online business configurations) is immense.

Re:frontend, you're absolutely correct. Magento's frontend is a sad state. MageOS is actively looking for ways to move Luma/Blank into a separate installable package so that devs are not forced to see it at every turn.

In terms of Headless, there are a bunch of different heads at this point, but none have yet solved what I call the "upgrade-safe theme" problem.

I like what we've done with Daffodil since it doesn't impose any of those limitations that come with themes, while still handling a lot of the complexity that devs face when building ecommerce stores.


Ignoring academia - what's the value for a business minded tech person to stay in the Magento world? I'm surely not the only M2 tech lead type who's jumped ship off to greener pastures.


Businesses will always have commerce problems, so you're pretty stable in terms of consistent income.

If you can make a name for yourself, you can do very well. Especially for Magento, I love the fact that I can make small change, deliver it as a composer package, and then (if the code is good enough) see the results on many brands within a few months.

It's hard to get that close-to-customer visibility anywhere else as most "cool" large scale FOSS projects are dev focused, which means your impact on the lay user is less obvious.


Thank you for calling them "JavaScript developers" instead of "front-end developers"; it's accurate.


I actively hire this way.


Do you give the newhires backpay?

Mostly joking but I don't think such a bonus is a crazy idea


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: