Socket exposes TrapDoor malware campaign targeting crypto and AI developers across global software ecosystems

A new cyberattack campaign is raising alarm bells across the crypto and artificial intelligence industries after security researchers uncovered a widespread supply chain attack aimed at developers.

The campaign, known as “TrapDoor,” has already spread through dozens of malicious software packages disguised as legitimate development tools.

According to cybersecurity platform Socket, the malware operation is actively targeting developers who work on cryptocurrency projects, decentralized finance applications, AI systems, and security infrastructure.

The attackers appear focused on stealing sensitive credentials, crypto wallet information, and confidential development data.

How the TrapDoor Attack Works

The attack relies on a tactic that has become increasingly common in recent years: poisoning developer package repositories with fake but convincing software tools.

Developers unknowingly install these packages during routine coding work, giving attackers access to sensitive systems.

Socket revealed that the campaign has already released more than 34 malicious packages and hundreds of related versions across multiple software ecosystems.

Researchers say the attackers continually upload new variants to avoid detection and remain active for longer periods.

The fake packages imitate useful developer utilities such as project setup tools, Solidity helpers, AI prompt engineering libraries, model-routing utilities, and blockchain build assistants.

Because many developers depend on open-source tools every day, the malicious packages blend naturally into normal workflows.

Crypto Wallets and Developer Secrets Under Attack

One of the most concerning aspects of the campaign is the range of sensitive information being targeted. Researchers say the malware attempts to steal:

• Crypto wallet credentials
• Secure Shell (SSH) keys
• Cloud service credentials
• GitHub access tokens
• API keys
• Browser extension data
• Development environment secrets

The malware specifically hunts for data connected to major crypto ecosystems and wallets, including Coinbase, Binance, Solana, Sui, Aptos, and MetaMask.

Even the privacy-focused Brave browser is reportedly among the targets.

AI Coding Assistants Are Being Manipulated

The campaign is not only attacking wallets and credentials.

Researchers say the attackers are also attempting to manipulate AI coding assistants such as Claude and Cursor.

Socket’s chief technology officer, Ahmad Nassri, explained that the malware injects hidden prompts and instructions designed to trick AI assistants into running fake security scans or workflows.

These manipulated workflows can then expose secrets stored on a developer’s machine.

This reflects a growing cybersecurity concern surrounding prompt injection attacks, where hidden instructions influence AI systems to behave in unintended ways.

npm, PyPI, and Rust Ecosystems Exploited

The malicious packages were discovered on several of the internet’s most important software repositories.

Suggested Readings (News continues below)

US Bitcoin ETF Investors Trigger Massive Withdrawals Across Funds in United States as Market Weakens After Six Day Outflow Streak

Israel Strike Injures Iran Supreme Leader Ali Khamenei as Secret Hideout Crisis Slows High-Stakes US Talks in Tehran

Enhanced Games Shock Las Vegas as Drug-Free Sprinter Fred Kerley Defeats Steroid-Fueled Rivals in Explosive 100m Final

Joey Rasdien Stuns Fans With Dramatic Weight Loss Transformation as South African Comedian Shows Off New Fitness Look Online in Johannesburg

Among the affected platforms are:

• npm for JavaScript and Node.js developers
• PyPI for Python developers and AI engineers
• Rust package repositories such as Crates, heavily used in blockchain and systems programming

These platforms form the backbone of modern software development.

Millions of developers worldwide rely on them daily, which makes them attractive targets for cybercriminals.

The attackers intentionally crafted package names to resemble legitimate development tools, increasing the chances that developers would install them without suspicion.

GitHub’s Role and Signs of AI-Assisted Malware Development

Researchers also traced elements of the campaign back to repositories hosted on GitHub.

According to Socket, the repositories contained patterns suggesting rapid iteration and partially automated malware creation.

Investigators observed security-themed templates, fake helper tools, prompt-injection instructions, and incomplete extraction modules mixed with functioning malicious code.

Security analysts believe some parts of the campaign may have been generated or accelerated with the help of AI tools, allowing attackers to produce and update malware variants far more quickly than traditional methods.

The revelations come shortly after GitHub disclosed unauthorized access to some of its internal repositories following the compromise of an employee device on May 20.

Why Supply Chain Attacks Are Becoming More Dangerous

Supply chain attacks have become one of the most effective cybercrime techniques because developers naturally trust popular software repositories and third-party libraries.

Unlike direct hacks, these attacks exploit trusted ecosystems.

A single compromised package can quietly spread to thousands of projects, businesses, and servers before being detected.

The crypto industry is especially vulnerable because developers often handle private keys, wallets, smart contracts, and large amounts of digital assets.

AI developers are also attractive targets due to their access to proprietary models, datasets, and cloud infrastructure.

Recent years have seen multiple incidents involving malicious npm and PyPI packages stealing credentials or planting backdoors inside applications.

Impact and Consequences

The consequences of the TrapDoor campaign could be severe for both individuals and companies.

If successful, attackers may gain access to:

• Cryptocurrency wallets holding valuable assets
• Private repositories containing sensitive code
• Cloud systems used for AI training and deployment
• Production infrastructure and deployment keys
• Corporate developer environments

For crypto projects, compromised credentials can lead to wallet theft, smart contract manipulation, or exchange breaches.

For AI companies, leaked models or cloud access could expose proprietary research and customer data.

The attack also highlights a broader problem facing the tech industry: developers increasingly rely on automated tools and AI assistants, creating new opportunities for hidden manipulation.

What’s Next?

Cybersecurity researchers expect the attackers behind TrapDoor to continue evolving their methods.

Because the campaign is already pushing new package versions rapidly, experts warn that additional malicious libraries could appear across more ecosystems.

Developers are now being advised to:

• Carefully verify package sources before installation
• Audit dependencies regularly
• Enable multi-factor authentication on developer accounts
• Avoid blindly trusting AI-generated workflows
• Monitor wallet and API access activity closely
• Use software composition analysis tools to detect malicious dependencies

Security teams are also expected to increase monitoring of open-source repositories as AI-assisted malware development becomes more common.

Summary

The TrapDoor campaign represents a dangerous evolution in supply chain cyberattacks, combining traditional malware tactics with AI manipulation techniques.

By disguising malicious software as ordinary developer tools, attackers are targeting the very people responsible for building crypto platforms and AI systems.

The operation’s focus on wallet theft, credential harvesting, and AI assistant hijacking shows how modern cybercriminals are adapting to emerging technologies.

With software ecosystems becoming increasingly interconnected, experts warn that supply chain attacks like this may continue to grow in scale and sophistication.

Bulleted Takeaways

• The “TrapDoor” malware campaign targets crypto and AI developers
• More than 34 malicious packages and hundreds of versions were discovered
• Attackers aim to steal wallets, SSH keys, API keys, cloud credentials, and GitHub tokens
• Major crypto platforms like Coinbase, Binance, Solana, Sui, Aptos, and MetaMask are targeted
• AI assistants such as Claude and Cursor are being manipulated through hidden prompts
• Malicious packages were found on npm, PyPI, and Rust package repositories
• Researchers believe parts of the operation may be AI-assisted
• Supply chain attacks are becoming one of the biggest threats facing developers and tech companies
• Security experts expect similar attacks to increase in the future