Critical RCE Flaws in n8n Workflow Platform Put Thousands of Instances at Risk

High-Severity Vulnerabilities (CVE-2026-1470, CVE-2026-0863) in n8n Platform Allow Remote Code Execution

CRITICAL
January 28, 2026
4m read
VulnerabilityCloud SecurityPatch Management

Related Entities

Organizations

JFrogShadowserver Foundation

Products & Tech

n8n

CVE Identifiers

CVE-2026-1470
CRITICAL
CVSS:9.9
CVE-2026-0863
HIGH
CVSS:8.5

Full Report

Executive Summary

Security researchers have disclosed two high-severity vulnerabilities in the n8n workflow automation platform, a popular open-source tool for connecting APIs and services. The most critical of these, CVE-2026-1470, is an eval injection flaw with a CVSS score of 9.9, which allows an authenticated attacker to break out of a security sandbox and achieve remote code execution (RCE). A second flaw, CVE-2026-0863 (CVSS 8.5), enables a similar RCE via the Python execution sandbox. Because n8n instances often store credentials and API keys for databases, cloud services, and internal applications, a successful exploit could grant an attacker a "skeleton key" to an organization's entire technology stack. The disclosure is especially alarming as data from the Shadowserver Foundation indicates that over 39,000 instances remain vulnerable to a separate critical RCE flaw, Ni8mare (CVE-2026-21858), disclosed weeks ago.


Vulnerability Details

The vulnerabilities, discovered by JFrog's security team, exploit weaknesses in the sandboxing mechanisms used by n8n to isolate user-provided code.

  • CVE-2026-1470 (CVSS 9.9 - Critical): This is an eval injection vulnerability in the n8n Expression sandbox. An authenticated attacker can craft a malicious JavaScript expression that, when evaluated, bypasses the sandbox's restrictions. This allows the attacker to execute arbitrary code on the main node of the n8n instance with the full permissions of the application process.

  • CVE-2026-0863 (CVSS 8.5 - High): This is also an eval injection flaw, but it targets the python-task-executor sandbox. An authenticated attacker can execute arbitrary Python code that escapes the sandbox and runs directly on the underlying operating system of the worker node.

These flaws highlight the inherent difficulty and risk of executing untrusted code, even within supposedly secure sandboxes. Subtle features of dynamic languages like JavaScript and Python can be abused to dismantle security boundaries.

Affected Systems

  • n8n instances are vulnerable if they are running versions prior to the patched releases. Users should update to the following versions or later:
    • For CVE-2026-1470: 1.123.17, 2.4.5, or 2.5.1
    • For CVE-2026-0863: 1.123.14, 2.3.5, or 2.4.2

Exploitation Status

There is no evidence of active exploitation of these two new CVEs. However, the public disclosure of the technical details means that threat actors will likely begin scanning for and attempting to exploit vulnerable instances. The fact that 39,000+ instances remain unpatched for the recent Ni8mare (CVE-2026-21858) vulnerability suggests a large attack surface of slow-to-patch systems.

Impact Assessment

A compromise of an n8n instance is a high-impact event. These platforms are central hubs of automation and are often configured with highly privileged credentials, API keys, and database connection strings for a wide array of services:

  • Cloud providers (AWS, GCP, Azure)
  • Databases (PostgreSQL, MySQL)
  • SaaS applications (Salesforce, Slack)
  • Internal IAM systems and LLM APIs

An attacker who gains RCE on an n8n server can likely exfiltrate these credentials, providing them with widespread access across an organization's infrastructure. This can lead to massive data breaches, financial theft, and compromise of core business systems.

Cyber Observables for Detection

  • Application Logs: Monitor n8n application logs for errors or warnings related to expression evaluation or sandbox execution. Unhandled exceptions or crashes could indicate an exploitation attempt.
  • Suspicious Workflows: Audit n8n workflows for any recently added or modified nodes that contain complex or obfuscated JavaScript or Python code.
  • Anomalous Processes: On the n8n server, monitor for the n8n process spawning unexpected child processes, such as shells (/bin/sh, cmd.exe) or network utilities (curl, wget).

Detection & Response

  1. Code Review: If possible, conduct a review of custom expressions and code nodes within your n8n workflows. Look for code that attempts to access the file system, environment variables, or make network connections in unexpected ways.

  2. Endpoint Detection (EDR): Deploy EDR on the server hosting n8n. Create rules to alert on the n8n process spawning shells or other suspicious binaries. Monitor for file creation in unusual directories.

  3. Network Monitoring: Monitor outbound network traffic from the n8n server. Alert on connections to unknown or suspicious IP addresses or domains, as this could be a sign of a reverse shell or data exfiltration.

Remediation Steps

  1. Update Immediately: The only effective remediation is to update all n8n instances to a patched version. This should be treated as a critical priority.

  2. Restrict Access: Ensure that access to the n8n user interface is restricted. It should not be publicly accessible. Use a firewall to limit access to trusted IP ranges and place it behind a reverse proxy with strong authentication (e.g., MFA).

  3. Credential Rotation: As a precautionary measure, consider rotating all credentials and secrets stored within n8n. If an instance was compromised before being patched, these secrets should be considered stolen.

Timeline of Events

1
January 28, 2026
This article was published

MITRE ATT&CK Mitigations

Updating to a patched version of n8n is the most effective way to remediate these vulnerabilities.

Mapped D3FEND Techniques:

While the internal sandboxes failed, running the entire n8n application within a hardened container (e.g., Docker) with strict resource limits and network policies can provide an additional layer of containment.

Mapped D3FEND Techniques:

Run the n8n service with the least privilege possible. This won't prevent RCE but can limit the immediate impact an attacker has on the underlying operating system.

Sources & References

Two High-Severity n8n Flaws Allow Authenticated Remote Code Execution
The Hacker News (thehackernews.com) January 28, 2026

Article Author

Jason Gomes

Jason Gomes

• Cybersecurity Practitioner

Cybersecurity professional with over 10 years of specialized experience in security operations, threat intelligence, incident response, and security automation. Expertise spans SOAR/XSOAR orchestration, threat intelligence platforms, SIEM/UEBA analytics, and building cyber fusion centers. Background includes technical enablement, solution architecture for enterprise and government clients, and implementing security automation workflows across IR, TIP, and SOC use cases.

Threat Intelligence & AnalysisSecurity Orchestration (SOAR/XSOAR)Incident Response & Digital ForensicsSecurity Operations Center (SOC)SIEM & Security AnalyticsCyber Fusion & Threat SharingSecurity Automation & IntegrationManaged Detection & Response (MDR)

Tags

n8nRCEsandbox escapeCVE-2026-1470CVE-2026-0863workflow automationvulnerability

📢 Share This Article

Help others stay informed about cybersecurity threats

Continue Reading