Description

Aix-DB through 1.2.4 renders markdown with raw HTML enabled into v-html bindings without sanitization, allowing stored cross-site scripting attacks. Attackers can inject malicious HTML and JavaScript through markdown content in chat responses, skill descriptions, or knowledge messages that execute in users' browsers when viewed.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "5.4",
  "pubDate": "2026-08-31T11:16:43.247Z",
  "pubdate": "2026-08-31T11:16:43.247Z",
  "executiveSummary": "The Aix-DB application, up to and including version 1.2.4, is susceptible to a stored Cross-Site Scripting (XSS) vulnerability.\nThis vulnerability stems from the improper handling of user-supplied markdown content, which is rendered using vulnerable v-html bindings without appropriate sanitization.\nThe vulnerability allows remote attackers to inject malicious HTML and JavaScript payloads into various application components, including chat responses, skill descriptions, and knowledge messages.\nBecause the payload is stored within the application, it triggers whenever a legitimate user views the compromised content in their browser.\nThe exploitation of this flaw can lead to unauthorized execution of arbitrary JavaScript, potentially resulting in session hijacking, unauthorized actions performed on behalf of the victim, information disclosure, and full compromise of the user's browser-side interaction with the application.\nThe risk is considered critical, as it requires no specialized user interaction beyond viewing the affected content within the application interface.",
  "technicalDetails": "The root cause of this vulnerability is the insecure implementation of the markdown rendering engine within Aix-DB. The application processes user-provided markdown input and renders it directly into the Document Object Model (DOM) using the v-html directive (a feature common in frameworks like Vue.js).\nThe v-html directive explicitly instructs the framework to treat the content as raw HTML, bypassing the framework's built-in automatic data binding protection that otherwise escapes HTML special characters.\nThe application lacks a server-side or client-side sanitization mechanism to filter out malicious scripts or dangerous HTML attributes before they are bound to the DOM.\nAn attacker can exploit this by injecting crafted markdown strings that include HTML tags such as <script>, <img> (with onerror attributes), or <a> (with javascript: pseudo-protocols) into any field that the application renders using the vulnerable v-html implementation.\nThe attack flow proceeds as follows: First, the attacker identifies an input vector where user-supplied data—such as a chat message, a skill description, or a knowledge base entry—is stored in the database. Second, the attacker submits a malicious payload containing an XSS vector. Third, the application stores this payload in its backend database without prior inspection or sanitization. Fourth, when a victim navigates to a part of the application that displays the attacker’s input, the application renders the malicious content into the DOM via the vulnerable v-html binding.\nUpon rendering, the victim's browser executes the injected JavaScript code in the context of the Aix-DB origin. This grants the attacker the ability to perform actions as the victim, bypass Same-Origin Policy (SOP) restrictions, steal session cookies via document.cookie access, or perform further client-side attacks.\nAffected versions include Aix-DB through 1.2.4. The vulnerability is persistent, meaning the payload remains active until it is manually removed from the database, ensuring that any user accessing the affected data is susceptible to the attack."
}