CVE-2026-6593: ComfyUI XSS Vulnerability

CVE-2026-6593 Overview

A Cross-Site Scripting (XSS) vulnerability has been identified in ComfyUI, a popular AI-powered image generation interface. This vulnerability affects the View Endpoint functionality within the server.py file in ComfyUI versions up to and including 0.13.0. An attacker can exploit this flaw remotely to inject malicious scripts into web pages viewed by other users, potentially leading to session hijacking, data theft, or unauthorized actions performed on behalf of authenticated users.

Critical Impact

Remote attackers can inject malicious scripts through the View Endpoint in ComfyUI's server.py, enabling client-side attacks against users of the affected application.

Affected Products

• ComfyUI versions up to 0.13.0
• ComfyUI View Endpoint component (server.py)

Discovery Timeline

• April 20, 2026 - CVE-2026-6593 published to NVD
• April 22, 2026 - Last updated in NVD database

Technical Details for CVE-2026-6593

Vulnerability Analysis

This vulnerability is classified as CWE-79 (Improper Neutralization of Input During Web Page Generation), commonly known as Cross-Site Scripting (XSS). The flaw resides in the View Endpoint handler within ComfyUI's server.py component. The application fails to properly sanitize or encode user-supplied input before including it in dynamically generated web pages.

When a user interacts with the vulnerable View Endpoint, an attacker can craft malicious requests containing JavaScript or HTML content. Because the application does not adequately validate or escape this input, the malicious content is rendered in the victim's browser context, allowing the attacker to execute arbitrary scripts.

The attack requires network access and some form of user interaction, typically clicking a malicious link or viewing attacker-controlled content through the ComfyUI interface. While the direct confidentiality impact is limited, the integrity of the user session and displayed content can be compromised.

Root Cause

The root cause of this vulnerability is insufficient input validation and output encoding in the View Endpoint functionality. The server.py component processes user-controlled data and reflects it back to the browser without proper sanitization, creating a reflected XSS attack vector. This is a common oversight in web applications where developers fail to treat all user input as potentially malicious.

Attack Vector

The vulnerability is exploitable remotely over the network. An attacker can craft a malicious URL or request payload targeting the View Endpoint in ComfyUI. When a victim user accesses this malicious content, the injected script executes in the context of the victim's browser session.

The exploitation mechanism involves:

1. An attacker identifying the vulnerable View Endpoint in server.py
2. Crafting a payload containing malicious JavaScript or HTML
3. Delivering the payload to potential victims via phishing, social engineering, or embedding in other content
4. The victim's browser executing the malicious script when processing the response from ComfyUI

A proof-of-concept demonstrating this vulnerability has been made publicly available. Technical details can be found in the GitHub Gist PoC Repository.

Detection Methods for CVE-2026-6593

Indicators of Compromise

• Unusual HTTP requests to the View Endpoint containing encoded script tags or event handlers
• Presence of JavaScript payloads in URL parameters or request bodies targeting server.py
• Client-side errors or unexpected script execution in browser developer consoles
• User reports of suspicious behavior or unexpected pop-ups when using ComfyUI

Detection Strategies

• Implement web application firewall (WAF) rules to detect and block XSS payload patterns in requests to the View Endpoint
• Monitor HTTP access logs for suspicious requests containing script injection patterns such as <script>, javascript:, or event handlers like onerror, onload
• Deploy Content Security Policy (CSP) headers to detect and report inline script execution attempts
• Utilize browser-based XSS auditing and monitoring tools to identify potential exploitation

Monitoring Recommendations

• Enable verbose logging for all requests to the ComfyUI View Endpoint
• Set up alerts for requests containing common XSS indicators or unusual URL encoding patterns
• Monitor for anomalous user session behavior that may indicate successful XSS exploitation
• Review application logs periodically for patterns consistent with XSS probing or exploitation attempts

How to Mitigate CVE-2026-6593

Immediate Actions Required

• Restrict network access to ComfyUI instances to trusted users and networks until a patch is available
• Implement input validation and output encoding on the View Endpoint as an interim measure
• Deploy a web application firewall with XSS protection rules in front of ComfyUI
• Educate users about the risks of clicking untrusted links while using ComfyUI

Patch Information

The vendor was contacted about this vulnerability but did not respond. As of the last update on April 22, 2026, no official patch has been released. Users should monitor the VulDB entry and the official ComfyUI repository for security updates. Consider upgrading to newer versions when they become available and verify they address this vulnerability.

Workarounds

• Place ComfyUI behind a reverse proxy that implements XSS filtering and input sanitization
• Implement Content Security Policy (CSP) headers to prevent inline script execution
• Restrict access to the View Endpoint using network-level controls or authentication mechanisms
• Consider disabling or limiting access to the View Endpoint functionality if not required for your use case
• Run ComfyUI in an isolated network environment to limit exposure to potential attackers