Malicious Open VSX Extension Used in $500K Crypto Heist
A deep dive into one of 2025’s most sophisticated open-source supply chain attacks
A malicious extension hosted on the Open VSX marketplace was used to steal $500,000 in cryptocurrency from a security-conscious developer. Disguised as a harmless Solidity syntax highlighter, the extension was quietly downloaded, activated, and used to gain remote access, steal credentials, and drain wallets.
Although the plugin has since been removed, the attackers have already begun uploading new variants—a troubling sign for the open-source ecosystem.
How the Scam Worked
The attackers deployed multiple strategies to make the extension appear legitimate:
Fake Popularity: Inflated download numbers to boost credibility
Search Manipulation: Used Open VSX’s ranking algorithm to appear above genuine tools
Rapid Re-Uploads: As soon as one plugin was taken down, clones were reuploaded under new names
Key Takeaway: Even in trusted developer ecosystems, malware can hide in plain sight. Verifying every extension before installation is no longer optional—it's essential.
The $500K Scam Hidden in a Code Highlighter
Malicious Open-Source Packages: A Growing Threat
Attacks leveraging public package repositories like npm, PyPI, and now Open VSX are on the rise. Despite heightened awareness and researcher vigilance, attackers continue to profit by exploiting open-source trust models.
Timeline of the Attack
June 2025
A Russian blockchain developer reached out to investigators after losing half a million dollars in cryptocurrency. The case was unusual:
The operating system had been freshly installed
Only trusted, widely used software had been downloaded
The developer was highly cautious with transactions
A full disk image was captured for forensic review
Discovery: Malware Masquerading as a Solidity Extension
The forensic analysis revealed a malicious file:
%userprofile%\.cursor\extensions\solidityai.solidity-1.0.9-universal\src\extension.js
This script fetched and executed a PowerShell payload from angelic[.]su
—clear evidence of malware.
The Fake Plugin in Cursor AI
The infected extension was a fake Solidity language support plugin on Open VSX, posing as a legitimate tool for Cursor AI (an AI-enhanced IDE built on VS Code).
What it claimed:
Syntax highlighting for Solidity smart contracts
Code optimization features
What it actually did:
Contained zero functional code
Copied its description from a legitimate plugin
Served only one purpose: download and execute malware
How the Victim Got Tricked
The developer searched for "solidity" in Cursor AI’s marketplace. The results:
Malicious extension: Ranked 4th, with 54K downloads
Legitimate extension: Ranked 8th, with 61K downloads
Why?
Open VSX ranks extensions by “relevance,” factoring in:
Download count
Recent updates
Ratings
Verification status
The attackers gamed the system:
Faked a recent update (June 15 vs. May 30)
Likely used bots to inflate downloads
Assuming the non-working extension was buggy, the victim left it installed—while malware ran silently in the background.
Infection Chain: From PowerShell to Full Takeover
Stage 1
The plugin downloaded 1.txt
(PowerShell) from angelic[.]su
.
Stage 2
If ScreenConnect wasn’t present, it downloaded 2.txt
, which:
Installed ScreenConnect from
lmfao[.]su
Connected to a C2 server:
relay.lmfao[.]su
Credential Theft & RAT Deployment
Next, the attackers executed three obfuscated VBS scripts:
a.vbs
,b.vbs
,m.vbs
Fetched more PowerShell payloads from
paste.ee
Downloaded hidden malware via archive[.]org image loaders
Payloads Deployed:
Quasar RAT: Remote access tool via
a.vbs
andb.vbs
Credential Stealer: Targeted crypto wallets, browsers, and email clients via
m.vbs
The stolen credentials led to the theft of $500,000 in digital assets.
The Attackers Rebrand and Relaunch
After the malicious plugin was removed on July 2, 2025, the attackers immediately:
Reuploaded a new extension named “solidity”—matching the legitimate name
Faked 2 million downloads to leapfrog legitimate results
Spoofed the developer name:
LegitFakejuanblancojuanbIanco
(with capital "i" instead of "l")
Now, the Cursor AI extension marketplace showed two nearly identical entries—one safe, one malicious.
Broader Campaign Discovered
Further investigation uncovered a wider campaign:
npm package
solsafe
: Dropped ScreenConnect viastaketree[.]net/1.txt
Other VS Code-style extensions:
solaibot
,among-eth
,blankebesxstnion
All used the same infection and obfuscation techniques
Common Attack Patterns
PowerShell scripts named
1.txt
,2.txt
Obfuscated VBS scripts calling
paste.ee
Payloads hidden in image files hosted on archive[.]org
Key Takeaways & Defense Tips
1. Verify Before You Trust
Check developer names closely
Review update dates and download spikes
If a plugin doesn’t work—inspect its code
2. Use Real Security Tools
Even basic antivirus could have blocked Quasar RAT and the stealer
3. Watch for Suspicious Activity
Monitor for unexpected PowerShell or ScreenConnect activity
4. Assume Repositories Can Be Dangerous
Open VSX, npm, PyPI, and others are frequent targets
Don't assume trust just because it’s open source
Final Thoughts
This incident proves what security professionals have warned for years: Open-source software can be weaponized, even in tools trusted by developers and engineers.
A single fake syntax highlighter cost one developer $500,000. The next victim could be a dev team, an exchange, or a DAO.
Trust nothing. Verify everything.
(Note: Malicious domains like
angelic[.]su
,lmfao[.]su
, and others are redacted for safety.)