This is a program that I've been championing within @nivenly over the past year, after we noticed that security vulnerabilities weren't being disclosed responsibly, and not enough research was going into the security of Fediverse software.
-
This is a program that I've been championing within @nivenly over the past year, after we noticed that security vulnerabilities weren't being disclosed responsibly, and not enough research was going into the security of Fediverse software.
You might remember my Pixelfed vulnerability from last year, where OAuth scopes weren't checked allowing for privilege escalation via the API (CVE-2024-25108), that was our very first test-case of this program.
I'm incredibly proud to be involved in launching the Fediverse Security Fund from Nivenly Foundation (a 501(c)4 not-for-profit cooperative)
@thisismissem@hachyderm.io what would buy-in from fediverse software look like?
NodeBB has its own bug bounty program that awards reporters directly, but if the FSF were to shoulder the grunt work of reporting (and act as a liaison between us and the reporter), we'd be happy to discuss covering the reward and associated costs, for reports that come from Nivenly directly.
-
J julian@community.nodebb.org shared this topic on
-
@julian you're still receiving the vulnerability reports directly with the Fediverse Security Fund; we pay *after* you've confirmed & patched.
I wasn't aware of your bug bounty program, but could list that alongside your project.
-
System shared this topic on
-
@julian you're still receiving the vulnerability reports directly with the Fediverse Security Fund; we pay *after* you've confirmed & patched.
I wasn't aware of your bug bounty program, but could list that alongside your project.
@thisismissem@hachyderm.io great. I'm thinking that for reports coming from Fediverse Security Fund directly, we'd cover the reward portion (the High (7.0 - 8.9) – $250 USD, Critical (9.0+) – $500 USD) part, either directly to the reporter or more likely through an in-kind donation back to the fund.
Also the fund may need a better acronym... FSF
-
@julian so the reports don't come from Nivenly, the reports come from researchers and contributors and go directly to you. Once you accept & fix, and publish the advisory, the researcher/contributor can come to us and we'll pay them for their responsible disclosure.
They could also still collect from your bounty program as well, so rather than them getting just $256 or $512 from your program, they could get $506 or $1012 in total, because they can claim both bounties (if your program allows it)
(I mean, it's better than Fediverse Security Bounty — FSB
) -
@julian so the reports don't come from Nivenly, the reports come from researchers and contributors and go directly to you. Once you accept & fix, and publish the advisory, the researcher/contributor can come to us and we'll pay them for their responsible disclosure.
They could also still collect from your bounty program as well, so rather than them getting just $256 or $512 from your program, they could get $506 or $1012 in total, because they can claim both bounties (if your program allows it)
(I mean, it's better than Fediverse Security Bounty — FSB
)@thisismissem@hachyderm.io ah understood. I didn't quite get how the fund worked, but it makes more sense now (and is much simpler—organizationally—for Nivenly!)
I don't think we'll add exclusions for security fund recipients
I would say, though, that one of the requirements has to be that the affected software accepts the vulnerability. Plenty of self-proclaimed "security researchers" have filed reports, and some go as far as to publish CVEs (against our own software!) without our permission.
Quite the opposite of responsible disclosure.
-
@julian yes, that's exactly what needs to happen. Like, it's CVE + the fix merged into the project. And we'll actually verify that before paying out. Definitely don't want those low quality reports for stuff that isn't actually a CVE
-
@julian yes, that's exactly what needs to happen. Like, it's CVE + the fix merged into the project. And we'll actually verify that before paying out. Definitely don't want those low quality reports for stuff that isn't actually a CVE
@julian we'll see how the fund goes, but we can always change the terms as necessary to get the right output, that's why this is an experiment.
-
@julian so the reports don't come from Nivenly, the reports come from researchers and contributors and go directly to you. Once you accept & fix, and publish the advisory, the researcher/contributor can come to us and we'll pay them for their responsible disclosure.
They could also still collect from your bounty program as well, so rather than them getting just $256 or $512 from your program, they could get $506 or $1012 in total, because they can claim both bounties (if your program allows it)
(I mean, it's better than Fediverse Security Bounty — FSB
)i feel like we missed an opportunity here @thisismissem by not choosing powers of two
love it @julian
-
i feel like we missed an opportunity here @thisismissem by not choosing powers of two
love it @julian
-
aaand aaah, TechCrunch have covered the announcement! Thanks @Sarahp!
A new security fund opens up to help protect the fediverse | TechCrunch
A new security fund aims to help apps in the fediverse — like Mastodon, Threads, and Pixelfed — to pay researchers for disclosing security bugs.
TechCrunch (techcrunch.com)
@thisismissem damn @Sarahp killing it with the fediverse coverage lately!
