Malicious code in xy-ai-chat (npm)
The npm package xy-ai-chat contains a Lit web component that silently exfiltrates all end-user chat input to a hardcoded attacker-controlled server (182.43.87.39) over plain HTTP with no TLS or configurability. Any site embedding this component routes user data to the attacker without consent or visibility.
- Disclosed
- Last updated
- Blast radius
- Any application embedding the xy-ai-chat component; all end-user chat input is exfiltrated to attacker-controlled server.
- Ecosystems
- Attack vectors
- Affected entities
- xy-ai-chatnpm package containing malicious Lit web component
The xy-ai-chat npm package ships a Lit web component with hardcoded plain-HTTP endpoints pointing to a bare IPv4 address (182.43.87.39) on ports 9050 and 9062. The component's bundled code in dist/index.esm.js contains fetch() calls that POST end-user chat content to http://182.43.87.39:9062/chat-ai-app-message/saveByUser with no encryption, domain-based routing, or user opt-in.\n\nThis is a silent-relay attack: the component's public API does not expose any configuration for these endpoints, meaning any website embedding xy-ai-chat automatically exfiltrates visitor chat input to the attacker's server. The package.json contains no homepage or publisher metadata identifying the operator of 182.43.87.39, and the data transmission uses plain HTTP, allowing any on-path observer to intercept the content.\n\nThe malicious behavior was identified by the OpenSSF malicious-packages project and reported via GitHub Advisory GHSA-vjg5-jmpf-7994.
Indicators of compromise
- Packages
- xy-ai-chat
- IPs
- 182.43.87.39
Remediation
- Immediately remove xy-ai-chat from all dependencies and applications
- Audit all applications that previously embedded xy-ai-chat for data exposure
- Review server logs for any POST requests to 182.43.87.39:9050 or 182.43.87.39:9062 to identify affected users
- Notify end-users whose chat input may have been exfiltrated
- Use npm audit or similar tooling to detect and prevent installation of this package
- Consider using alternative, vetted web component libraries with transparent data handling
Sources
- GitHub Advisory GHSA-vjg5-jmpf-7994 · GitHub Advisory Database
Cite this entry
"Malicious code in xy-ai-chat (npm)." supplychainattack.org, Supply Chain Attack Incident Catalog. Disclosed July 27, 2026; last updated July 27, 2026. https://supplychainattack.org/incident/malicious-code-in-xy-ai-chat-npm-1lhm2i
Suggest a correction
Found an error or have a newer source? Corrections to factual errors take priority over new entries.
Related incidents
- activecritical
Malware in edu-npm-dependency-chain-demo
Malware discovered in the npm package edu-npm-dependency-chain-demo. Systems with this package installed are considered fully compromised and require immediate remediation.
npmCompromised package - containedcritical
Malware in roblox-api-client
Malware was discovered in the npm package roblox-api-client, resulting in full system compromise for any computer with the package installed or running. The advisory recommends immediate removal of the package and rotation of all secrets and keys from a different computer.
npmCompromised package - containedcritical
Malware in jextic-eclib
Malware was discovered in the npm package jextic-eclib, resulting in full system compromise of any computer with the package installed or running. All secrets and keys on affected systems should be rotated immediately from a different computer.
npmCompromised package - activecritical
Malware in @thone33/analytics-injector
Malware discovered in the npm package @thone33/analytics-injector. Systems with this package installed are considered fully compromised and require immediate remediation.
npmCompromised package