Trapdoor malware campaign targets crypto developers' tools

TLDR

A coordinated malware campaign called TrapDoor is compromising popular developer packages to steal wallet keys and credentials from crypto focused developers.

1. TrapDoor has infected at least 34 packages and 384 versions across npm, PyPI, and Crates.io, silently exfiltrating crypto wallets, SSH keys, and cloud credentials.
2. The campaign specifically targets crypto and blockchain environments, including wallets like MetaMask and Solana tools, and even tries to hijack AI coding assistants to leak secrets.
3. Developers should urgently review dependencies, rotate exposed keys, harden AI and build environments, and watch for new indicators of compromise as this supply chain attack is still evolving.

Deep Dive

1. What TrapDoor Does

Security firm Socket detailed TrapDoor as a large scale supply chain attack that injects malware into open source packages across npm, PyPI, and Crates.io, spanning 34 packages and 384 versions published in waves starting around 22 May 2026. These packages pose as generic developer tools, then run malicious code during normal install, import, or build steps, so no extra action is needed from the victim.

Once executed, TrapDoor scans machines for high value data such as crypto wallet files, SSH keys, GitHub tokens, browser stored credentials, and cloud API keys, then sends them to attacker controlled servers. Researchers report more than 30 malicious packages and over 300 affected versions distributed in a short window, confirming that this is a coordinated operation, not isolated uploads.

2. Why It Matters For Crypto

TrapDoor is tuned for crypto and Web3 targets rather than generic enterprise environments. Reports highlight searches for wallet related files and credentials tied to platforms like MetaMask, Coinbase and Binance wallets, Solana, Sui and Aptos tooling, and Brave browser profiles that often hold extension data.

The attackers chose developers because they often control project treasuries, deployment keys, and CI credentials. A single compromised dev box can eventually expose protocol admin keys, multisig signers, or production RPC and cloud access, which can cascade into exchange, DeFi, or bridge level incidents.

On top of that, TrapDoor attempts to abuse AI coding assistants by dropping directive files such as special config or markdown instructions that push tools like Claude or Cursor to run scans and export sensitive data. This reflects a shift where AI assisted workflows themselves become part of the attack surface.

3. How To Reduce Risk

For crypto developers, immediate steps include:

1. Audit recent dependency additions or upgrades, especially from lesser known maintainers, and remove any packages later flagged as malicious.
2. Assume potential credential exposure and rotate wallet keys, SSH keys, and cloud or GitHub tokens that were accessible from affected machines.
3. Treat AI assistants and their connector configs as untrusted input, review any custom rules or MCP connectors, and prefer scanners that inspect config and package metadata without executing code, as described in a recent open source tool release focused on supply chain threats.

Longer term, teams can reduce blast radius by isolating signing and treasury operations from general dev laptops, enforcing least privilege on CI and cloud roles, and pinning dependencies to vetted versions rather than auto updating.

Conclusion

TrapDoor is a concrete example of how crypto risk is shifting from end user phishing to deep software supply chain compromise of developer tooling and AI assisted workflows. If you build or maintain crypto infrastructure, the key is to treat every layer of your toolchain, including AI configs, as a security surface and to rotate any credentials that might already be exposed while the community continues to uncover affected packages.