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
cortex9-nuxt-dot-segment-middleware-bypass-model — Offline educational model of CVE-2026-119200 (Nuxt dot-segment middleware bypass) | Kitploit
ツール/GitHubGitHub/federiconbaez/cortex9-nuxt-dot-segment-middleware-bypass-model
Vulnerability AnalysisCode AnalysisWeb Application ExploitationWeb SecurityPapers & ResearchLearning & EducationLabs & Practice
GitHub
federiconbaez/cortex9-nuxt-dot-segment-middleware-bypass-model

cortex9-nuxt-dot-segment-middleware-bypass-model

Offline educational model of CVE-2026-119200 (Nuxt dot-segment middleware bypass)

リポジトリを見る
11日前未レビュー

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。

    1.1k
  • 2

    ghostlock-app

    yukonga/ghostlock-app

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

    599
  • 3

    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。

    394
  • 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。

    321
  • 5

    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。

    277
  • 6

    Aliens_eye

    arxhr007/aliens_eye

    ウェブ全体で7,700以上のソーシャルメディア上の身元を発見・マッピングするAI搭載OSINTエンジン

    257

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 →
共有
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。

cortex9-nuxt-dot-segment-middleware-bypass-model

Offline model of CVE-2026-119200 (GHSA-9v7c-q64q-vv33): in Nuxt 4.6.0-4.6.2 a page request with . / .. segments (also spelled %2e) renders the page the path resolves to, while server/middleware and the server route rules see the unresolved path. The model compares the behaviour before (pre) and after (post) the fix commit 45463f8 (released in [email protected] / @nuxt/[email protected]). Educational, deterministic, Python standard library only. No Nuxt, h3, Nitro or browser code is imported or run and no socket is opened.

Spanish: README.es.md. Exercises: PRACTICE.md. Limits: LIMITATIONS.md. Sources: SOURCES.md.

What is modelled

diff:N is line N of the saved fix diff (SOURCES.md). Ported code is marked in src/nuxt_model/model.py; everything else is labelled RECONSTRUCTED.

  1. event.path (reconstructed): one percent-decoding pass that keeps %2F (diff:17 comment; the advisory gives /%2e/admin -> /./admin).
  2. Server middleware: the advisory's example, event.path.startsWith('/admin') -> 401 (config mw).
  3. Route rules (config rules): /dashboard headers, /old -> /new (both from the diff test at diff:122-135) and the advisory's /admin/** redirect. Lookup through normalizeRouteRulePath: pre = decode + fold (diff:210, old:40); post = decodeRoutePath(resolveRoutePathDotSegments(path)) (diff:211, new:42; runtime copy diff:230, 235). resolveRoutePathDotSegments is ported line by line from diff:154-176 (packages/nuxt/src/core/utils/index.ts new:60-82).
  4. Renderer toRequestEvent (packages/nitro-server/src/runtime/utils/event.ts): post only, 404 when the path portion of event.path matches DOT_SEGMENT_RE = /(?:^|\/)\.{1,2}(?:\/|$)/ (diff:15 = new:12; diff:28-31 = new:271-274).
  5. Page router (reconstructed): encodeEventPath (encodeURI, diff:12-14 constants), WHATWG URL dot-segment resolution ("as url would resolve them", diff:24-25), then one decode, case-insensitive match, optional trailing slash, against the pages / /admin /admin/secret /dashboard /old /new /login.

Configs: bare (only the renderer, like the diff test at diff:42-100), mw, rules. Labels: BYPASS_DOT (protected page rendered, middleware skipped, dot segments present), RULE_SKIPPED_DOT (the rule keyed on the rendered page was not applied), DOT_404 (post-fix rejection), AUTH_401, REDIRECT_307, NOT_FOUND_404, URL_EQ / URL_DIFF (bare: does the page URL name the same path as event.path?), CASE_ONLY_NOT_CVE (/ADMIN: case-insensitive page router, mentioned in the advisory's workaround notes, not fixed by 45463f8 and not this CVE), OK, BAD_INPUT.

Differential evidence

CaseprepostDiff lines
g01 /%2e/admin (mw)BYPASS_DOT: event.path /./admin, page /adminDOT_404diff:28-31
g03 /x/%2e%2e/admin (mw)BYPASS_DOTDOT_404diff:28-31
g04 /admin (mw)AUTH_401AUTH_401none (control)
g05 /%2e/admin/secret (rules)RULE_SKIPPED_DOT (no redirect)REDIRECT_307 to /logindiff:154-176, 211
d15 /./dashboard (rules)RULE_SKIPPED_DOT (no header)DOT_404, header rule now resolveddiff:129-130, 211
d01-d09 (bare)URL_DIFFDOT_404test table diff:76-88
d10-d14 (bare)URL_EQURL_EQtest table diff:90-100
a01 /ADMIN (mw)CASE_ONLY_NOT_CVECASE_ONLY_NOT_CVEnot changed by the fix

Fixed sweep (results/sweep.csv): 1884 targets (/ + 1-3 tokens from admin ADMIN x old '' . .. %2e .%2e %2e. %2e%2e %252e) x configs mw, rules. pre: 192 BYPASS_DOT, 260 RULE_SKIPPED_DOT. post: 0 of either; 2658 DOT_404. A sweep is a fixed grid, not a proof.

Run

python3 tests/run.py                                 # 26 tests, a skip fails
python3 scripts/derive_expected.py | diff - fixtures/expected.tsv
python3 scripts/derive_expected.py sweep | diff - results/sweep.csv
PYTHONPATH=src python3 -m nuxt_model results fixtures/cases.tsv /tmp/r && diff -r results /tmp/r
PYTHONPATH=src python3 -m nuxt_model case pre /%2e/admin mw

scripts/derive_expected.py does not import src/; it recomputes every expectation with different algorithms (RFC 3986 remove_dot_segments, urllib.parse.quote, a byte-loop decoder, regex rule matching).

A model is not a finding: see LIMITATIONS.md.

License: Apache-2.0.

ツールをダウンロード