MalwarePublished 15 Jul 2026

ai-explain (npm) · Malicious Code

Affects ai-explainSource GCVE Database

Malicious package : remove ai-explain and rotate any secrets on hosts that installed it.

Data as of 15 Jul 2026, 13:05 UTC

Classification

Malicious package

Published to npm · OSV MAL-2026-10628

What to do

Remove ai-explain from every project, lockfile and image that pulls it in. Treat any host that installed it as compromised: rotate the credentials, tokens and keys it could read, from a clean machine.

Affected software

Product	Affected	Fixed in
npmai-explain	Affected =0.3.4	Fixed in–

Check your own hosts for these packages with threat-finder (opens in a new tab), our free open-source scanner.

Description

The npm package ai-explain version 0.3.4 contains malicious code that collects host environment information including OS username, hostname, current working directory, platform, architecture, and Node.js version. This data is sent to a hardcoded external endpoint during installation and each CLI invocation without user disclosure. The package impersonates a legitimate project to lure users. This behavior enables tracking of developer or CI machines by the attacker.

Impact

Users installing or running ai-explain@0.3.4 unknowingly expose detailed host environment data to an attacker-controlled server. This enables attacker tracking of developer or CI environments, potentially facilitating follow-on targeting or reconnaissance. There is no indication of direct code execution or data corruption, but privacy and operational security are compromised.

Defensive guidance

Users should avoid installing or running ai-explain@0.3.4. Remove the package from any environments where it is present. Monitor for any suspicious network connections to the domain livekit-agents.xyz and block it if possible. Check vendor advisories or npm security notices for updates regarding this package.

Summary

The ai-explain@0.3.4 npm package includes a postinstall script and CLI code that silently harvests host identifiers such as the OS username, hostname, current working directory, platform, architecture, and Node.js version. This information is transmitted via HTTPS POST requests to a hardcoded domain (livekit-agents.xyz) controlled by the attacker. The package falsely advertises itself as an AI code explanation tool but exports only a trivial greet function. The malicious domain and dependency mimic the legitimate LiveKit project to deceive users. This results in undisclosed host tracking on every install and CLI execution.

References

npmjs.com/package/ai-explain/v/0.3.4 (opens in a new tab)

Community reviews

Loading reviews…

What to do

Affected software

Description

AI analysisMachine-generated, updated 15 Jul 2026, 13:05 UTC

Technical details

References

Community reviews