air-security-asdfoingeroi3234dsfds.webflow.io Fake GitHub Repo Delivers SolanaCry Malware via AI Agents
Article Content
- •SolanaCry malware targets AI coding agents via a fake GitHub repository.
- •The attack exploits trust in AI agents, allowing malware installation without user intervention.
- •Malware includes legitimate NetSupport components, enabling remote access to infected systems.
AIR Research Lab has identified SolanaCry, a malware targeting AI coding agents through a fraudulent GitHub repository masquerading as a Solana starter kit. This repository, created in November 2025, ranked first in search results and instructed AI agents to execute a PowerShell command that installs remote-access malware on Windows machines. The malware, which includes legitimate components of NetSupport Manager, allows attackers to control infected systems and access cryptocurrency wallets. The attack exploits the trust developers place in AI agents, making it a novel threat vector. Researchers noted similarities to previous campaigns, particularly one from October 2025 that also targeted cryptocurrency traders. The repository had 286 stars at the time of discovery, indicating significant visibility. The attack method involves fileless execution, hiding its presence from users. Current status indicates ongoing risks as the malware remains undetected by many users.
Ask AI about this cluster
Answers cite the sources they use
Timeline
More articles in this cluster (4)
Following this threat?
Track NetSupport RAT in your own feed — you're alerted when they show up in new reporting, leak sites or exploitation.
Free account · no card needed
Common questions
How does SolanaCry infect systems?
What are the implications for developers?
What should organizations do to protect against this threat?
Continue Reading
Malware Campaign Exploits Hugging Face for Crypto Mining A cryptomining and backdoor campaign has resurfaced, exploiting Hugging Face model repositories to hijack GPUs. Researchers at Pluto Security identified 31 malicious repositories across seven accounts, with the most popular model downloaded over 28,000 times. The attack leverages the 'trust_remote_code' option…
Warden Stealer Malware Targets AI Agents for Data Theft Warden Stealer, a sophisticated malware-as-a-service, is actively targeting AI agents like Claude, Codex, Grok, and Cursor to steal sensitive developer data. This Rust-based infostealer collects configuration files, tokens, and conversation histories, posing a significant threat to organizations using AI tools. The…