Skip to content
KitploitKITPLOIT
ツールエクスプロイトブログ
Log in
提出
ツールエクスプロイトブログ
提出

ハッキング、侵入テスト、サイバーセキュリティツールをあなたのセキュリティアーセナルに!

Kitploitはハッキング、サイバーセキュリティ、ペネトレーションテストのツールディレクトリです。最新のプロジェクトアップデートを見つけて、脆弱性の発見、システム分析、テストの自動化、セキュリティの強化を行いましょう。

フィードお問い合わせプライバシー© 2026 Kitploit

ツールディレクトリ

カテゴリ

すべてのカテゴリを見る
Loading categories
omnicheck_CVE-2026-98432 — Non-destructive checker for CVE-2026-98432 (Linux kernel TCP use-after-free): patch status, exposed listeners and mitigation verification. By OMNIVERSE. | Kitploit
ツール/GitHubGitHub/omniversetr/omnicheck_cve-2026-98432
Defensive ToolsPacket Sniffing & AnalysisReconnaissanceVulnerability ScannersNetwork MappingPort ScanningVulnerability AnalysisConfiguration Auditing
GitHubomniversetr/omnicheck_cve-2026-98432

omnicheck_CVE-2026-98432

Non-destructive checker for CVE-2026-98432 (Linux kernel TCP use-after-free): patch status, exposed listeners and mitigation verification. By OMNIVERSE.

1910日前未レビュー
共有
リポジトリを見るウェブサイト
要求された言語のコンテンツは利用できません。英語版を表示しています。

omnicheck-98432

English | Türkçe

Non-destructive verification tool for CVE-2026-98432 (Linux kernel IPv4 TCP use-after-free)

Developed by OMNIVERSE Siber Güvenlik ve Bilişim A.Ş.


What it does

omnicheck-98432 determines the exposure of a Linux system to CVE-2026-98432 without changing anything on the system and without triggering the vulnerability:

  • Checks whether the kernel is patched using distribution evidence (rpm/deb changelog, kpatch, Livepatch, KernelCare) and the upstream fixed-version table
  • Lists non-loopback IPv4 TCP and MPTCP listeners, including dual-stack [::] sockets that accept IPv4 and container network namespaces
  • Checks net.mptcp.enabled and net.ipv4.tcp_migrate_req
  • Verifies whether the temporary mitigation (dropping TCP packets with IPv4 options) is loaded as a drop rule and persistent
  • Checks a whole server list over SSH in one run with your own keys (fleet), with a summary table and CSV
  • Optionally proves external reachability and that the mitigation actually works, at packet level

It reports a one-line verdict, an exit code and a JSON report.

What it does not do

It does not trigger the race condition, does not send SYN retransmissions, does not corrupt memory, does not attempt privilege escalation and does not change any setting. It sends at most three single SYN packets per port.

Quick start

python3 omnicheck-98432.py selftest                          # SELFTEST_OK
sudo python3 omnicheck-98432.py local --namespaces           # on the target server
python3 omnicheck-98432.py fleet --hosts-file servers.txt --out-dir reports --csv summary.csv   # all servers over SSH

Sample local run on an Ubuntu 24.04 server with the temporary mitigation applied (host name and IPs anonymized):

omnicheck-98432 local mode output

Verdict line:

  SONUÇ   VULNERABLE_MITIGATED   Zafiyetli çekirdek; geçici önlem aktif, dağıtım yaması çıkınca güncelleyip yeniden başlatın

The report is printed in Turkish; verdict codes (PATCHED, VULNERABLE_MITIGATED, ...), JSON field names, --help and source comments are in English.

Modes

ModeRuns onProduces
localTarget Linux server (sudo)Patch status, listeners, MPTCP, mitigation, one-line verdict
probeScanning machine (root + scapy)Reachability of the trigger packet class and MPTCP negotiation; targets as HOST[:PORTS] or CIDR[:PORTS], sent in paced batches
captureTarget server (sudo)Passive record of incoming SYNs (IPv4 options, MPTCP)
auditScanning machine (root + scapy + SSH)local + capture + probe in one run, packet-correlated evidence
fleetManagement machine (SSH keys)local on every server in a list over SSH; one JSON per server, summary table, CSV
combineAnywhereMerges multiple JSON reports into one table; newest report per server wins; optional CSV
selftestAnywhereOffline parser tests

Checking many servers over SSH (fleet)

python3 omnicheck-98432.py fleet --hosts-file servers.txt --out-dir reports --csv summary.csv

servers.txt holds one SSH destination per line (web-01, [email protected], [email protected]:2222, # comments; see examples/). The script is streamed over stdin, nothing is copied or installed on the targets. fleet never prompts (BatchMode): it needs key-based login and passwordless sudo, or a root account. Servers without a result are reported as UNREACHABLE, SUDO_REQUIRED, PYTHON_MISSING, PYTHON_TOO_OLD or REMOTE_ERROR and count as exit code 3. Use --remote-python python3.11 for RHEL 8 / Rocky 8, --parallel for concurrency (default 8).

The scanning machine's own addresses are skipped by probe, so a block that contains it can be scanned as is.

Where to run probe from

probe works from anywhere, but its IP-options result is only conclusive when the path to the target does not filter IP options. Many internet paths (ISPs, CGNAT, cloud edges) silently drop packets with IP options; from such a vantage point TCP_REACHABLE_IPOPT_NO_REPLY means "the packet did not get through", not "the host mitigation works". Run probe from the same network as the targets (DMZ, management network, jump host) or pair it with capture on the target (audit does this automatically) to get a definitive answer.

sudo python3 omnicheck-98432.py probe 10.0.0.0/24:22,443 --output dmz.json    # one /24, two ports
sudo python3 omnicheck-98432.py probe --targets-file scope.txt --ports 22      # file: HOST or CIDR per line

Endpoints are deduplicated by resolved IP and port; up to 8192 endpoints per run, 50 packets/s by default (--rate).

Exit codes

CodeMeaning
0Patched or not affected
1Vulnerable and unprotected
2Vulnerable, temporary mitigation active
3Undetermined, inconclusive or tool error

fleet and combine return the worst server: 1 > 3 > 2 > 0. No servers at all is 3, never 0.

Verdicts

VerdictMeaning
PATCHEDThe running kernel's package changelog or an applied livepatch names the CVE, or a non-distribution kernel is at a fixed upstream version
NOT_AFFECTEDKernel older than 4.4 (e.g. RHEL 7)
VULNERABLE_NO_LISTENERVulnerable kernel, no non-loopback IPv4 TCP listener
VULNERABLE_MITIGATEDVulnerable kernel, IPv4-options drop rule active
VULNERABLE_EXPOSEDVulnerable kernel, listeners present, no mitigation
INCONCLUSIVENot enough evidence: unparseable kernel release, a distribution kernel at an upstream fix level without package evidence, or listeners could not be listed
VULNERABLE_MITIGATED_VERIFIED(audit) Mitigation active; on the same endpoint the plain SYN was answered and the SYN with IPv4 options arrived intact and was not answered
VULNERABLE_TRIGGER_PATH_OPEN(audit) SYN with IPv4 options was answered; trigger path is reachable
CRITICAL_MPTCP_REMOTE(audit) Vulnerable kernel and MPTCP negotiation accepted from outside

Requirements

  • Target server: Linux, Python 3.8+, sudo. No additional libraries.
  • Scanning machine (probe, audit): Python 3.8+, root, scapy (pip3 install scapy), OpenSSH client.
  • Management machine (fleet): Python 3.8+, OpenSSH client and keys for the targets; no root, no scapy.

The default python3 on RHEL 8 / Rocky 8 is 3.6; use python3.9 or python3.11 on those systems.

Documentation

Full guide covering all modes, parameters, verdicts and scenarios:

  • English: docs/omnicheck-98432-user-guide.pdf
  • Türkçe: docs/omnicheck-98432-kullanim-kilavuzu.pdf

Responsible use

Use this tool only on systems you are authorized in writing to test. The probe and audit modes send raw packets and may trigger IDS/IPS alerts; notify the relevant security team before testing. This tool is not an exploit and does not prove exploitability.

License

MIT

Contact

OMNIVERSE Siber Güvenlik ve Bilişim A.Ş. · omni-verse.io · [email protected]

ツールをダウンロード

人気

すべて見る →

コミュニティで最も使われているツールを見つけましょう。

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 究極のOSINTツールキット:調査グラフ、信頼度スコアリング、エイリアスファミリー、マップ、タイムライン、レポート。49個のモジュール、400以上のサイト、APIキー不要。

    916
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    GhostLock ワンタップ実行アプリ(CVE-2026-43499)

    558
  • 3

    redforge

    cybersecurity91/redforge

    ブラウザ側のレッドチームツールキット。偵察、ペイロード生成、認証情報攻撃、暗号、エンコーディング、エンゲージメントレポート向けの102のオフラインツールを備えています。

    370
  • 4

    ARTEX

    baarai/artex

    プランナー/ワーカーエージェント、資産グラフと探索グラフ、MITMトラフィック記録、脆弱性の再テストを備えたLLMマルチエージェント自律ペネトレーションテストプラットフォーム。

    360
  • 5

    looking-glass

    anonghosty/looking-glass

    ShadowserverレポートをMongoDBに取り込み、インタラクティブなマップ、分析、調査結果、アドバイザリ、エグゼクティブダッシュボードとして再生するCSIRTプラットフォーム。

    317
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    AIネイティブなサイバーセキュリティのためのアクションシステム。意図がガバナンスされた実行に変わり、証拠が運用記憶となり、すべての操作が次の操作を改善する。

    289

すべてのツールを探索

ツールコレクションを閲覧

すべてのツールを見る →