In 30 seconds
On November 15, 2026, a researcher using the handle v4bel announced a project called NoisyTCP, pointing followers to a public GitHub repository and posting the disclosure on X at the same moment.github.comgithub.com · rssOpen original sourceSee in source listv4belx.com · v4bel · xOpen original sourceSee in source list
According to the researcher, NoisyTCP demonstrates CVE-2026-99288, a flaw in the Linux kernel's TCP stack, the core code that handles network connections. The claim is stark: an attacker could corrupt a machine's memory over the network without any login or authentication, and the defect has allegedly been sitting in the kernel for about 11 years. The advice that came with it was blunt: patch it.v4belx.com · v4bel · xOpen original sourceSee in source list
What happened
The story starts with two synchronized artifacts dated November 15, 2026: a GitHub repository for NoisyTCP under the V4bel account, and an X post from v4bel announcing it. Both carry the same timestamp, so this looks like a coordinated self-disclosure rather than a leak or a response to someone else's findings.github.comgithub.com · rssOpen original sourceSee in source listv4belx.com · v4bel · xOpen original sourceSee in source list
In the post, the researcher describes the work as Linux kernel TCP pre-auth remote corruption, files it under the identifier CVE-2026-99288, says the defect has been latent for 11 years, and closes with a direct instruction to patch. The repository is named as the home of the project, but its contents are not included in the material available here.v4belx.com · v4bel · xOpen original sourceSee in source list
Why it matters
If the claim holds, the exposure would be unusually broad. TCP is the protocol behind nearly every network service, so almost any Linux host that talks to a network could be in scope, and pre-auth means an attacker would need no credentials, no account, and no user interaction, only a path to the target's network stack.v4belx.com · v4bel · xOpen original sourceSee in source list
There is also a practical gap that shapes how much this matters right now: the announcement urges patching, but nothing in the available evidence shows a released fix, an affected version list, or a vendor advisory. Until those arrive, defenders can see the warning but cannot yet act on it directly.v4belx.com · v4bel · xOpen original sourceSee in source listgithub.comgithub.com · rssOpen original sourceSee in source list
What to do
The researcher's stated guidance is simple: patch the kernel as soon as a fix reaches you. That is the entire recommendation in the announcement, with no workaround, mitigation, or configuration change offered alongside it.v4belx.com · v4bel · xOpen original sourceSee in source list
Given that no fix is confirmed in the current evidence, a reasonable reading of the situation is to track the identifier CVE-2026-99288 through your distribution's security advisories and treat kernel updates in the coming days as time-sensitive.v4belx.com · v4bel · xOpen original sourceSee in source list
Confirmed and still uncertain
What is established at this point: the announcement exists, it comes from v4bel, it names the project NoisyTCP, it uses the identifier CVE-2026-99288, it claims pre-auth remote corruption in the Linux kernel TCP stack, and it urges patching.v4belx.com · v4bel · xOpen original sourceSee in source listgithub.comgithub.com · rssOpen original sourceSee in source list
What is not established: independent confirmation of the flaw, a kernel maintainer or distribution acknowledging it, affected kernel versions, a proof of concept, exploit conditions, or a shipped fix. The 11-year latency figure is the researcher's own estimate, presented in quotation marks, and the repository contents are not part of the evidence here.v4belx.com · v4bel · xOpen original sourceSee in source listgithub.comgithub.com · rssOpen original sourceSee in source list
What people are saying
The researcher framed the disclosure as a turning point: The era of KVM escapes is fading, and the era of TCP remote exploits has begun. KVM, or Kernel-based Virtual Machine, is Linux's built-in virtualization layer, and the argument is that attackers will shift from breaking out of virtual machines to breaking into hosts directly through the TCP stack. That is the researcher's interpretation, not a demonstrated trend.v4belx.com · v4bel · xOpen original sourceSee in source list
Evolution
The timeline is short so far. Both the repository and the announcement post appeared on November 15, 2026, and no follow-up, correction, retraction, or third-party coverage is present in the current evidence. This is the opening move of the story, not its development.github.comgithub.com · rssOpen original sourceSee in source listv4belx.com · v4bel · xOpen original sourceSee in source list
Reach
1 independent groups, 1 updates, and 6 engagements.
Technical detail
The technical substance, as claimed: a pre-authentication remote corruption bug in the Linux kernel's TCP stack, meaning an unauthenticated remote party could alter kernel memory, filed under CVE-2026-99288. The researcher puts the figure of 11 years in quotation marks, which reads as an approximation or a borrowed number rather than a measured result.v4belx.com · v4bel · xOpen original sourceSee in source list
What the announcement itself does not contain is as telling: no affected kernel versions, no word on whether a service must be listening to be reachable, no proof-of-concept description, and no patch reference. The repository URL is the designated primary artifact, but its contents are not visible in this evidence batch.v4belx.com · v4bel · xOpen original sourceSee in source listgithub.comgithub.com · rssOpen original sourceSee in source list
Sources and evidence
The evidence base is two items from the same origin: the NoisyTCP GitHub repository, marked as primary, and v4bel's X post, marked as discussion, both timestamped November 15, 2026. The repository entry carries a URL but no body text or media.github.comgithub.com · rssOpen original sourceSee in source listv4belx.com · v4bel · xOpen original sourceSee in source list
Read this as a single-source, self-reported disclosure. The existence of the flaw, its age, and its exploitability all remain allegations until a kernel maintainer, a distribution, or a CVE authority corroborates them, so the right stance is high attention with low confidence for now.v4belx.com · v4bel · xOpen original sourceSee in source list
Sources
- github.comgithub.com · rss
- v4belx.com · v4bel · x



