Improper Authorization in ChatGPTNextWeb MCP Server Configuration

A vulnerability has been found in ChatGPTNextWeb NextChat up to 2.16.1. Affected is the function addMcpServer of the file app/mcp/actions.ts. The manipulation leads to improper authorization. Remote exploitation of the attack is possible. The exploit has been disclosed to the public and may be used. The project was informed of the problem early through an issue report but has not responded yet.

Casky was already ahead

This CVE exploits attack patterns that Casky's 261matched skills have been training on — long before this vulnerability was disclosed. Claude's reasoning model maps these techniques to MITRE ATT&CK, so practitioners who ran these skills have already seen the threat behaviour in their findings.

Analysis

ChatGPTNextWeb NextChat versions up to 2.16.1 contain an improper authorization vulnerability in the addMcpServer function (app/mcp/actions.ts) that allows remote attackers to manipulate server configurations without proper access controls. This vulnerability affects users and deployments of NextChat who rely on MCP (Model Context Protocol) server integration, potentially exposing sensitive configurations and allowing unauthorized modifications to critical infrastructure components. The vulnerability has been publicly disclosed and is currently exploitable, making it a priority for organizations running vulnerable versions of NextChat.

Casky's 261 mapped security skills leverage Claude AI's extended reasoning to detect the authorization bypass patterns associated with MITRE techniques TA0004 (Privilege Escalation) and TA0006 (Credential Access). Practitioners using Casky will identify suspicious activity patterns including: unauthorized API calls to the addMcpServer endpoint lacking proper session validation or role-based access controls (TA0004), configuration changes executed by accounts without administrative privileges, and potential lateral movement attempts through compromised MCP server configurations (TA0006). The platform's skill mapping to CWE-266 (Improper Privilege Management) and CWE-285 (Improper Authorization) will surface abnormal function call sequences, missing authentication checks before sensitive operations, and deviations from expected authorization workflows—enabling practitioners to detect and respond to exploitation attempts before successful compromise occurs.

261 Casky skills cover this attack pattern

These skills use Claude AI's reasoning model to surface findings in the same attack categories as CVE-2026-7644.

analyzing-cloud-storage-access-patterns

cloud security · low

Run

analyzing-kubernetes-audit-logs

container security · low

Run

analyzing-malicious-url-with-urlscan

phishing defense · medium

Run

MITRE ATT&CK Techniques

TA0004 TA0006

Practise detecting this in the Casky Playground

Run Claude AI skills against real targets and see findings like those behind CVE-2026-7644 — before the next CVE drops.

Join the Playground Waitlist →

X Instagram LinkedIn