sales@hackhalt.com

Home / Threats / CVE-2026-130121

AI Puffer - Security Vulnerability (CVE-2026-130121)

CVE-2026-130121

The AI Puffer - MCP Server for Claude & ChatGPT, with AI Chatbot plugin for WordPress is vulnerable to Stored Cross-Site Scripting via AI-generated comment replies in all versions up to, and including, 2.5.28 due to insufficient sanitization of the AI model's output in the insert_comment_reply_logic() function, which saves replies with wp_insert_comment() and bypasses the kses filtering WordPress normally applies to comment content. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page, by submitting a comment that leads the AI model to reproduce an HTML payload in its automated reply. This is only exploitable when an administrator has configured a Comment Auto-Reply automation task with 'Action on reply' set to 'Approve' (the default is 'Hold for moderation'), and the attacker's comment has been approved.

CVE-2026-130121

HIGH CVSS 7.2 Published 2026-11-18 Updated 2026-11-18
AI Risk High (80/100) Active Exploit: No strong signal Published Exploit: No public exploit references Priority: P2 Urgent
Severity Band HIGH
CVSS Vector CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:L/A:N
Affected Components 2
Reference Links 5
AI Risk Engine High (80/100)
Exploitability High
Active Exploitation No strong signal
Published Exploit Status No public exploit references

Threat Timeline

  1. 2026-11-18 CVE published and first recorded in the threat feed.
  2. 2026-11-18 Record updated with latest vulnerability metadata.
  3. 2026-11-20 AI technical context refreshed for mitigation and impact guidance.
  4. Now Monitoring for follow-up changes, linked references, and new related CVEs.

AI Context

Machine-generated threat intelligence

AI Updated 1 day ago

AI enriched 1 day ago (2026-11-20 22:35 UTC)

Technical Summary

The AI Puffer - MCP Server for Claude & ChatGPT, with AI Chatbot plugin for WordPress is vulnerable to Stored Cross-Site Scripting via AI-generated comment replies in all versions up to, and including, 2.5.28 due to insufficient sanitization of the AI model's output in the insert_comment_reply_logic() function, which saves replies with wp_insert_comment() and bypasses the kses filtering WordPress normally applies to comment content. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that will execute…

Potential Impact

Severity is HIGH (CVSS 7.2). Depending on deployment context, affected components may be exposed to unauthorized actions or data integrity risk.

Exploitability Assessment

Exploitability is assessed as High based on remote code execution potential, low-bar exploit prerequisites.

Primary risk drivers: remote code execution potential, low-bar exploit prerequisites

Mitigation Recommendations

Validate affected product versions, prioritize patching, and monitor references for vendor remediation guidance. If immediate patching is not possible, apply compensating controls and limit exposure of vulnerable surfaces.

Detection & Monitoring

Track authentication anomalies, unexpected file writes, and suspicious plugin API activity around affected components.

Business Impact Lens

AI risk score 80/100 (High, High) with priority P2 Urgent. Prioritize remediation where affected components process customer data, admin sessions, or Internet-exposed workflows.

Affected Products

AI Puffer PLUGIN · ai-puffer Affected: <= 2.5.28 Fixed version not specified
Gpt3 Ai Content Generator PLUGIN · gpt3-ai-content-generator Affected: <= 2.5.28 Fixed version not specified

References

What to do about this

Patch promptly The consequence is significant even where an attacker needs a foothold first.

Fix it

Update AI Puffer to the latest available version.

Where the vendor has not published a fix yet, disable the plugin until they do.

Cannot update today? Request / payload inspection (WAF) in Hack Halt ProSecure blocks this class of attack before WordPress processes the request.

Inspect request parameters and bodies for injection, traversal and file-inclusion payloads, and refuse the request rather than logging it.

Compare Hack Halt plans Start free

Not sure if you are affected?

Scan your site to see whether it runs AI Puffer, and at which version.

Free, no installation, no account. It reads what your site already publishes, then tells you which plugins carry known vulnerabilities and what to do about each one.

Scan my site free See what Hack Halt does

Am I affected by this vulnerability?

You are affected if your site runs AI Puffer. No account is needed to exploit this, so a public site is reachable by anyone.

How do I fix it?

Update AI Puffer to the latest available version. Where the vendor has not yet published a fix, remove or disable the plugin until they do. Until you can update, Request / payload inspection (WAF) in Hack Halt ProSecure contains this class of attack at the request layer.

What can an attacker actually do?

This is a Cross-Site Scripting (XSS) flaw. An attacker who exploits it can run script in a visitor's browser, stealing sessions or planting an administrator account.