---
title: "SafeDep Blog"
url: "https://safedep.io/blog"
---

# SafeDep Blog

- [Malicious npm Package Turns Hermes Agent into a Backdoor (2026-11-19)](https://safedep.io/blog/malicious-npm-leadstack-crm-connect-hermes-agent-backdoor.md): The malicious npm package @leadstack/crm-connect installed Hermes Agent on Linux servers, loaded it with keys and instructions from the attacker's server, and connected it to the attacker's Telegram bot as a boot-time systemd service.
- [Microsoft agent-framework Compromised Twice on PyPI (2026-11-18)](https://safedep.io/blog/microsoft-agent-framework-pypi-compromise.md): Versions 1.26.1 and 1.26.2 of agent-framework, the Microsoft Agent Framework metapackage on PyPI, shipped a credential stealer that runs at every Python start. SafeDep flagged both. The second release went up 33 minutes after the first, with the payload renamed and GitHub Actions OIDC theft added.
- [Malicious wppconnect Release Steals Secrets and WhatsApp Sessions (2026-11-16)](https://safedep.io/blog/wppconnect-npm-whatsapp-session-stealer.md): Version 2.4.1 of @wppconnect-team/wppconnect and version 2.10.82 of @wppconnect/server shipped a credential stealer that runs on import or server start. It sends environment variables, .env files and WhatsApp session tokens to a Telegram bot. The repository history was then rewritten to hide the code.
- [The Rules That Moved safedep.io from Astro to Next.js 16 (2026-11-13)](https://safedep.io/blog/safedep-io-astro-to-nextjs-migration.md): Lint before code, useEffect banned, complexity capped at 15, clicks 2.4× faster, and an interactive guide to choosing your Next.js caching setup.
- [wfx: GitHub Actions Credential Stealer Hits 8,000 Repositories (2026-11-12)](https://safedep.io/blog/github-actions-wfx-credential-stealer.md): Since November 10, 2026, an attacker has used stolen GitHub tokens to push a credential stealer into more than 8,000 repositories. It runs once in GitHub Actions or on npm install, sends secrets and cloud credentials to the attacker, and leaves cloud backdoors. Then the attacker reverts the commit.
- [Compromised bee-js and Swarm Desktop Steal Swarm Wallet Keys (2026-11-09)](https://safedep.io/blog/ethersphere-bee-js-npm-compromise.md): Version 13.3.0 of @ethersphere/bee-js, the JavaScript client for Swarm, added a postinstall hook that steals the Swarm node wallet key and installs a remote-access backdoor. Swarm Desktop 0.56.2 shipped the same loader.
- [Backdoored PyPI Package fragment-api-py 14.4.2 Drains TON Wallets on Startup (2026-11-10)](https://safedep.io/blog/fragment-api-py-pypi-ton-wallet-drainer.md): Version 14.4.2 of fragment-api-py, a PyPI client for Telegram Fragment, sends all TON and USDT from the user wallet to an attacker wallet when the client opens. The release is the old 13.0.0 code with a backdoor and a new version number.
- [testmetest npm Packages Carry a Miasma-Family Worm (2026-10-22)](https://safedep.io/blog/testmetest-npm-worm-miasma-variant.md): 19 npm packages carry a Miasma-family worm that steals credentials, spreads to other packages and installs a wiper. The latest wave plants instructions in GitHub repositories that tell AI coding agents to run the worm.
- [Binance Hack Lessons: Sandbox the Take-Home Assignment (2026-11-04)](https://safedep.io/blog/binance-hack-lessons.md): A fake take-home assignment replaced a Homebrew binary on a Binance work laptop. SafeDep rebuilt that setup step as a benign lab and ran it with and without the pmg sandbox. The post gives the commands, the real output and the limits of the sandbox.
- [Binance Hack: How Attackers Stole $6.34 Billion Through Developer Laptops (2026-11-03)](https://safedep.io/blog/binance-hack-developer-supply-chain-attack-path.md): The October 29, 2026 Binance hack started on developer laptops. A tampered Homebrew install and a fake job interview gave a North Korean group its first access. A short breakdown of the attack path.
- [antd-img-crop GitHub Repository Hijacked to Deliver OverLord Stealer (2026-11-03)](https://safedep.io/blog/antd-img-crop-forged-commit-overlord-stealer.md): A forged commit force-pushed to the antd-img-crop GitHub repository runs a dropper on npm run build that installs a cross-platform stealer and remote-access agent called OverLord. The npm package is clean.
- [October 2026 Security Incident Review: PHP, Binance, Cursor, LiteLLM and More (2026-11-01)](https://safedep.io/blog/october-2026-security-incident-review.md): October was Cybersecurity Awareness Month. It was also the month PHP, Railway, OpenCode, Binance, Polymarket, Cursor, LiteLLM and eight more made the news for a security incident. A short recap of what happened to each one.
- [eMudhra CA Key Compromise: Response Guide for Developers (2026-10-22)](https://safedep.io/blog/emudhra-ca-key-compromise.md): Anyone can compute eight eMudhra intermediate CA private keys from public certificates. Browsers block the compromised intermediates, but CI runners, containers, JVMs and Go binaries can still trust forged certificates. Test and update affected systems, and decide whether to rotate credentials.
- [Compromised LiteLLM 1.107.0 on PyPI Installs a Credential Stealer and Backdoor (2026-10-21)](https://safedep.io/blog/litellm-pypi-compromise-credential-stealer-backdoor.md): Every wheel of LiteLLM 1.107.0 on PyPI installs a credential stealer and a remote code execution backdoor. Both run each time Python starts. The code is not in the upstream GitHub source.
- [Forged Commits Push a preinstall Dropper into hugo-paper and mevn-cli (2026-10-20)](https://safedep.io/blog/hugo-paper-mevn-cli-forged-commit-preinstall-dropper.md): An attacker pushed forged commits to the default branches of hugo-paper and mevn-cli. Running npm install or bun install in either repository downloads and runs a second-stage script from buildwithknexus[.]xyz.
- [MALFEX Moves Its npm Loader into a Native Windows DLL (2026-10-13)](https://safedep.io/blog/malfex-imgdecode-npm-native-addon-loader.md): The MALFEX package imgdecode contains a native Windows loader. When loaded from a console, it attempts to download, decrypt, and inject shellcode into dllhost.exe.
- [ASOS App Hack: Attackers Used Push Notifications to Send a Ransom Note (2026-10-09)](https://safedep.io/blog/asos-push-notification-extortion-snowflake-claim.md): Attackers got into the platforms ASOS uses to message customers and pushed an extortion note to phones. Here is what is confirmed, what is only claimed, and what customers should do.
- [tensorlake 0.5.144 npm Compromise Ships Mini Shai-Hulud (2026-10-08)](https://safedep.io/blog/tensorlake-npm-compromise-mini-shai-hulud.md): tensorlake@0.5.144 on npm ran a Mini Shai-Hulud worm at install. A maintainer account committed the payload through the GitHub web interface, and the official release workflow published it.
- [42 Malicious RubyGems Open a Reverse Shell at Install Time (2026-10-05)](https://safedep.io/blog/rubygems-crypto-gems-install-time-reverse-shell.md): One RubyGems account published 42 gems that run code during gem install. On a developer workstation, the gems open a reverse shell or download a second stage. They do nothing on CI runners and sandboxes.
- [An Attacker Hijacked an AI Coding Assistant to Spread a Worm (2026-10-02)](https://safedep.io/blog/ai-coding-assistant-hijack-shai-hulud.md): An attacker took over a live AI coding assistant session, got it to recommend a poisoned package, and used the stolen tokens to spread the Shai-Hulud worm across about 100 internal repositories.
- [PolinRider Switches to Ethereum C2 in 30+ Repositories (2026-09-30)](https://safedep.io/blog/polinrider-ethereum-c2-github-repositories.md): A malicious pull request against oxc led SafeDep to a larger campaign. The PolinRider loader family now reads its C2 servers from Ethereum transactions. SafeDep confirmed 35 GitHub repositories that still carry the loader, two operator wallets, and 11 C2 servers.
- [DirtyBlanket: Fake Express Packages on npm Spread a Linux Worm (2026-09-29)](https://safedep.io/blog/dirtyblanket-express-impersonation-npm.md): Nine fake Express and React packages on npm run a Linux worm at install time. It installs a Tor backdoor and spreads through SSH, AUR packages, and npm tokens.
- [Detecting Compromised AI Coding Agents with Jev and Gryph (2026-09-25)](https://safedep.io/blog/ai-agent-security-jev-gryph.md): I checked every action of my own Claude Code agent against a profile of how I work and a set of org policies, using Jev. It caught 14 of 14 attacks for $0.15 per 1,000 events.
- [Mini Shai-Hulud Is Still Infecting GitHub Repositories (2026-09-24)](https://safedep.io/blog/mini-shai-hulud-reinfection-github-repositories.md): The May 2026 Mini Shai-Hulud worm is still infecting new GitHub repositories. Hijacked actions-cool/issues-helper tags planted Claude Code and VS Code hooks in six popular repositories between 20 and 24 September.
- [MemTensor npm and PyPI Packages Hit by a Go Worm (2026-09-23)](https://safedep.io/blog/memtensor-sckit-worm-npm-pypi.md): An attacker used a Go worm to steal CI publish tokens from MemTensor and ship malicious MemOS packages to npm and PyPI. See how it works, with code and indicators of compromise.
- [sckit: A New Go Implant Framework for Supply Chain Worms (2026-09-23)](https://safedep.io/blog/sckit-go-implant-framework.md): sckit is a Go implant framework that steals developer and CI credentials and carries templates to spread through npm, PyPI, and GitHub Actions. Static analysis and indicators of compromise.
- [OpenAI Agents Turned RubyGems Into a Scraping Proxy (2026-09-18)](https://safedep.io/blog/openai-agents-rubygems-attack.md): Between May and July 2026, a swarm of AI agents published over 3,000 packages to RubyGems. The gems abused RubyDoc.info documentation builds to run a crawler on someone else's servers, then shipped the scraped data back to the registry as new gems and webhook URLs. Three independent researchers traced the campaign to OpenAI agents four months later.
- [Why Does an npm Math Library Need an Encrypted Loader? (2026-09-18)](https://safedep.io/blog/mathmain-encrypted-loader.md): A math solver leads to encrypted code in npm packages. Follow the loader, the trigger matrix that unlocks it, the remote access payload, and the full indicators.
- [The Agentic IDE Extension Blind Spot (2026-09-11)](https://safedep.io/blog/cursor-extension-import-versions.md): Cursor can install most of the same extensions you had in Visual Studio Code, but not the same versions. Its Import VS Code Configuration step sends only the extension name, never the version. It installs the newest version its own store has, which for most extensions is whatever Open VSX calls latest.
- [Deep-Live-Cam Supply Chain Attack: Technical Analysis (2026-09-09)](https://safedep.io/blog/deep-live-cam-supply-chain-attack.md): A malicious dependency in Deep-Live-Cam loads a clipboard hijacker. We trace the installation trigger, Telegraph delivery, and Windows and macOS persistence.
- [Introducing SafeDep Threat Intel (2026-09-08)](https://safedep.io/blog/introducing-safedep-threat-intel.md): SafeDep Threat Intel gives SOC and cyber defense teams the malicious package intelligence behind SafeDep's platform, to query or to push into the tools they already run.
- [A malicious npm package hidden three dependencies deep: the ulid-xyz delivery chain (2026-09-01)](https://safedep.io/blog/ulid-xyz-transitive-dependency-delivery-chain.md): ulid-xyz is a remote access trojan we reported to OSV as MAL-2026-6672. It reached developers as a third-level transitive dependency of ioredis-xyz, a copy of the real Redis client planted in 28 attacker-created GitHub repositories. The implant matches the DPRK-linked MicrosoftSystem64 cluster.
- [Mini Shai-Hulud Strikes Again: openapi-react-query-codegen (2026-08-29)](https://safedep.io/blog/mini-shai-hulud-openapi-react-query-codegen-compromised.md): An attacker exploited a flawed GitHub Actions workflow to publish 10 malicious versions of @7nohe/openapi-react-query-codegen (671K downloads/month) via npm OIDC trusted publishing. The payload uses a binding.gyp code execution technique via node-gyp and a multi-layer obfuscated JS loader to download and execute a Bun binary.
- [From YouTube Ad to Root: How a Fake TradingView Installer Delivers a macOS Stealer (2026-08-21)](https://safedep.io/blog/youtube-ad-fake-tradingview-macos-stealer.md): A YouTube video ad impersonating TradingView delivered a fake .pkg installer. Inside: a self-healing LaunchAgent, an AES-encrypted V8 bytecode payload, and a Node.js MITM proxy trusted by a rogue root CA.
- [Malicious Rust Crate arrayref Runs a Build-Time Payload (2026-08-20)](https://safedep.io/blog/arrayref-proc-macro1-rust-build-time-malware.md): A compromised release of the popular Rust crate arrayref pulled in a typosquatted proc-macro1 whose build script downloads and runs a remote binary at compile time. The malicious versions are now removed from crates.io.
- [npm Bin Entry Harvesting: A Dependency Confusion Blind Spot (2026-08-14)](https://safedep.io/blog/google-dep-confusion-bin-harvesting.md): 21 malicious npm packages targeted Google by squatting CLI binary names from scoped packages, not package names. The technique exploits a structural gap that standard dependency confusion defenses do not cover.
- [Baileys npm forks farm WhatsApp Channel followers (2026-08-10)](https://safedep.io/blog/malicious-baileys-npm-whatsapp-campaign.md): A growing npm campaign uses forks of the Baileys WhatsApp library to make developers' own accounts follow attacker channels, inflating follower counts and injecting advertising.
- [keyv and cacheable npm compromise: 400+ packages (2026-08-04)](https://safedep.io/blog/keyv-npm-supply-chain-compromise.md): An npm worm published 2,234 poisoned versions across 444 package names. Its install script stole credentials and ran a command when GitHub revoked a token.
- [Joyfill npm Packages Compromised with Blockchain C2 Loader (2026-07-29)](https://safedep.io/blog/joyfill-npm-blockchain-c2-supply-chain.md): Malicious beta versions of @joyfill/components and @joyfill/layouts published on July 28, 2026 carried the PolinRider blockchain dead drop loader inside their production bundles. The Tron-to-BSC C2 chain, campaign markers, and XOR key rotation match the infrastructure documented in the earlier astro.config.mjs investigation.
- [mrmustard PyPI Package Trojanized to Steal Credentials (2026-07-24)](https://safedep.io/blog/mrmustard-malicious-pypi-package.md): Version 0.7.4 of Xanadu's mrmustard quantum computing library shipped to PyPI with no matching GitHub release. It carries a 258-line credential stealer that harvests SSH keys, AWS and Kubernetes credentials on import and installs three separate persistence hooks, published through a compromised maintainer account after CI tokens were stolen.
- [@copilot-mcp/apex: A macOS Infostealer Re-Published on npm After Takedown (2026-07-22)](https://safedep.io/blog/malicious-copilot-mcp-apex-npm-macos-infostealer.md): The npm security team removed the original @apexfdn/apex package for malicious code, and the operator re-published the same postinstall macOS infostealer as @copilot-mcp/apex about 11 hours later. It targets Web3 founders and developers, stealing browser data, 20+ cryptocurrency wallets, and SSH/AWS/Kubernetes credentials via osascript and curl|zsh, and phones home to a live command-and-control server every 60 seconds. npm has since removed the re-published package too, but the GitHub binaries and C2 infrastructure remain live.
- [AsyncAPI Packages Compromised with Miasma RAT (2026-07-14)](https://safedep.io/blog/asyncapi-generator-supply-chain-attack-miasma-rat.md): Four @asyncapi npm packages were published with obfuscated malware on July 14, 2026 via compromised CI workflows. The payload downloads Miasma RAT, a credential stealer targeting browsers, SSH keys, npm tokens, and cryptocurrency wallets. @asyncapi/specs alone has 2.7M weekly downloads.
- [Official jscrambler npm Package Compromised Across Multiple Releases (2026-07-11)](https://safedep.io/blog/jscrambler-npm-supply-chain-compromise.md): The official jscrambler npm package (60K monthly downloads) was trojanized starting at 8.14.0 through an npm account or CI compromise. The attacker republished the same Rust infostealer across five releases in three hours, including 19 minutes after a clean remediation release, and pivoted from a preinstall hook to running the dropper from the module code itself to slip past install-script scanners.
- [nodemon-sudo: an npm Backdoor With No Install Script (2026-07-09)](https://safedep.io/blog/malicious-nodemon-sudo-tslint-conf-npm-backdoor.md): nodemon-sudo copies the real nodemon byte for byte, adds nothing malicious to its own code, and injects one extra dependency, tslint-conf, a repackaged pino logger carrying a backdoor. It has no install hook. The payload fires only at runtime, one dependency hop away from the package a developer installed.
- [@marketfront: 25 npm Packages Reuse a Known Lure (2026-07-02)](https://safedep.io/blog/marketfront-dependency-confusion-campaign.md): On July 1, 2026, npm user marketfront batch-published 25 packages carrying the same README lure SafeDep has tracked across four earlier accounts (mr.4nd3r50n, pik-libs, t-in-one, emcd-vue): "Internal package — Platform Engineering Team". The marker is a reusable dependency-confusion template. It stayed constant across every wave while the payload behind it evolved from a process.env beacon into a credential-file harvester with an RC4-hidden C2.
- [The Polymarket Trap: A Fake Arbitrage Bot, Ten npm Accounts, and Four Ways to Deliver an Infostealer (2026-06-30)](https://safedep.io/blog/defi-infostealer-fake-arbitrage-bot-npm.md): A GitHub repository posing as a Polymarket arbitrage bot accumulated 53 forks before anyone flagged the malicious npm package buried in its dependencies. Behind that repository: ten coordinated npm accounts, 30 packages, and four delivery techniques including one that uses npm itself as the payload server. The second-stage credential harvester reads cryptocurrency wallets, browser credentials, AWS keys, SSH keys, and password managers.
- [Miasma Worm Infects Multiple LeoPlatform npm Packages (2026-06-25)](https://safedep.io/blog/miasma-worm-hits-leoplatform-20-npm-packages.md): A Miasma worm variant compromised a single maintainer account and used it to publish infected versions of 20 LeoPlatform npm packages within a 3-second window. The worm also pushed weaponized GitHub Actions workflows to at least three repositories. The payload is the same Bun-based credential stealer and self-propagating worm documented in our Miasma source code analysis.
- [The wshu.net npm Campaign Delivers a Multi-Stage Infostealer (2026-06-23)](https://safedep.io/blog/wshu-net-npm-credential-stealer-campaign.md): One actor seeded 15 npm packages across 13 throwaway scopes in a single morning, each shipping a ~270KB obfuscated downloader behind a postinstall hook. The downloader pulls a Rust infostealer from GitHub Releases that drains cryptocurrency wallets, browser credentials, cloud tokens, and SSH keys. Amazon Inspector tracks the same cluster as Operation Friday Harvest.
- [MYRA: A Full Linux RAT Distributed via npm (2026-06-21)](https://safedep.io/blog/malicious-apintergrationpost-npm-myra-rat.md): The npm package apintergrationpost is a red team RAT called MYRA with native C rootkit, triple persistence, fileless execution, live screen streaming, and process masquerade. This analysis documents its full capability set so defenders can detect and respond if it is misused.
- [@withgoogle/stitch-sdk: Scope Squat Harvests Developer Credentials (2026-06-20)](https://safedep.io/blog/withgoogle-stitch-sdk-scope-squat-credential-harvester.md): A malicious npm package squats the @withgoogle scope to impersonate Google Stitch, silently harvesting credentials from Claude Code, git, GitHub CLI, SSH keys, npm, and Docker on install.
- [Five npm Packages That Hide a Windows Binary Dropper (2026-06-17)](https://safedep.io/blog/procwire-npm-windows-dropper-campaign.md): Five npm packages published in a 12-minute burst split a Windows binary dropper across a fake utility toolkit. The loader hides in a preinstall hook, decodes its C2 from a helper package, and fetches a payload from catbox.moe.
- [Mastra npm Scope Takeover: 143 Packages Drop a RAT (2026-06-17)](https://safedep.io/blog/mastra-npm-scope-takeover-supply-chain-attack.md): An attacker republished 143 @mastra packages, including @mastra/core, each with one injected dependency: easy-day-js, a dayjs clone whose install hook downloads and runs a remote access trojan.
- [astro.config.mjs Supply Chain Attack via Blockchain C2 (2026-06-12)](https://safedep.io/blog/astro-config-blockchain-c2-supply-chain.md): An obfuscated IIFE hidden in astro.config.mjs fires at every build, beacons an HTTP C2, and pulls staged commands from a Tron-to-BSC blockchain dead drop.
- [Miasma Worm: Most Infected GitHub Repositories Are Still Live (2026-06-11)](https://safedep.io/blog/miasma-worm-still-infected-github-repos.md): Eight days after the Miasma worm forged a credential stealer into public GitHub repositories, most are still serving it. A re-scan of the published victim list plus a fresh code-search sweep found 123 repositories across 56 accounts still carrying the live 4.3 MB payload on 665 branches. The big names cleaned up. The long tail did not.
- [Inside the Miasma Software Supply Chain Attack Toolkit (2026-06-09)](https://safedep.io/blog/inside-the-miasma-supply-chain-attack-toolkit.md): The Miasma worm source code appeared on GitHub through compromised developer accounts. The codebase is a full supply chain attack toolkit with credential exfiltration across AWS, Azure, GCP, and Kubernetes, registry infection for npm, PyPI, and RubyGems via stolen tokens and OIDC trusted publishing, GitHub Actions tag hijacking, AI coding tool configuration poisoning, and a GitHub commit search based C2 infrastructure.
- [Configuration Files That Run Code: Supply Chain Security Blindspot (2026-06-06)](https://safedep.io/blog/config-files-that-run-code.md): Editor and package-manager configuration files auto-execute commands when a developer opens a folder or installs dependencies. The Miasma worm wired one dropper into seven of them across Claude Code, Gemini, Cursor, VS Code, npm, Composer, and Bundler. Opening a cloned repository is no longer safe.
- [Miasma Worm Targets AI Coding Agents via GitHub Repositories (2026-06-05)](https://safedep.io/blog/miasma-worm-ai-coding-agent-config-injection.md): A Miasma worm variant injects a 4.3 MB dropper into GitHub repositories across multiple maintainers, wiring it to auto-run through Claude Code, Gemini, Cursor, and VS Code configuration files. No npm package is published. The trigger is cloning a repository and opening it in an AI coding agent, a shift from the campaign's earlier node-gyp install-time execution.
- [Axios Typosquats Deliver the Epsilon Stealer (2026-06-02)](https://safedep.io/blog/malicious-faster-axios-npm-epsilon-stealer.md): Two axios typosquats on npm, turbo-axios and faster-axios, form a campaign delivering Epsilon Stealer through a four-stage chain. The Electron infostealer grabs browser credentials, cryptocurrency wallets, Discord tokens, and opens a persistent WebSocket RAT.
- [Mini Shai-Hulud "Miasma: The Spreading Blight" Hits @redhat-cloud-services: Multiple Packages at Risk (2026-06-01)](https://safedep.io/blog/redhat-cloud-services-hit-by-mini-shai-hulud-npm-worm.md): The attacker compromised the @redhat-cloud-services GitHub Actions OIDC trusted publisher to ship patch-client@4.0.4 with a Mini Shai-Hulud worm. The same publisher controls 32 packages across the scope. The payload harvests cloud, CI, and registry credentials and self-propagates through stolen tokens.
- [Inside MicrosoftSystem64: A Supply Chain RAT Exfiltrating to HuggingFace (2026-05-28)](https://safedep.io/blog/microsoftsystem64-binary-payload-analysis.md): Deep technical analysis of MicrosoftSystem64, an 81 MB Node.js SEA binary deployed via malicious npm packages. This RAT steals browser credentials, 80+ cryptocurrency wallet extensions, Telegram sessions, SSH keys, and screenshots, exfiltrating everything to HuggingFace datasets while maintaining persistence across Windows, macOS, and Linux.
- [183 npm Packages Target Cloud and Finance via oob.moika.tech (2026-05-28)](https://safedep.io/blog/oob-moika-tech-dependency-confusion-campaign.md): Two npm accounts published 164 malicious packages at version 99.99.99 targeting a cloud platform and a financial institution. Both campaigns share identical payload code, the same C2 endpoint, and the same hardcoded secret — strong evidence of a single actor. Updated June 1 with a fourth account (emcd-vue, 3+ packages) impersonating the EMCD cryptocurrency exchange, now with WaCk/Jscrambler obfuscation, home-directory persistence, and a FUSION_ second-stage protocol.
- [141 npm Packages Abuse Registry as Adware Hosting (2026-05-27)](https://safedep.io/blog/malicious-npm-terminal3airport-proxy-adware-spam.md): npm account terminal3airport published 141 packages containing a web proxy unblocker disguised as tutoring websites. The packages load popunder ads, external monetization scripts, and Google Analytics tracking, using npm as free static file hosting. An auto-publish shell script automates the spam at scale.
- [forge-jsxy: 22 Versions of an Actively Developed npm RAT (2026-05-26)](https://safedep.io/blog/malicious-forge-jsxy-npm-rat-evolution.md): forge-jsxy picked up where the taken-down forge-jsx left off, publishing 22 versions over 22 days. Each release added new capabilities: cryptocurrency wallet scanning, Chromium extension theft, WebRTC data channels, and relay-driven auto-upgrades. This post traces that evolution version by version.
- [Megalodon: Mass GitHub Repository Backdooring via CI Workflows (2026-05-21)](https://safedep.io/blog/megalodon-mass-github-repo-backdooring-ci-workflows.md): Over 5,700 malicious commits were pushed to GitHub repositories on May 18, 2026, replacing GitHub Actions workflows with base64-encoded secret exfiltration payloads. The "megalodon" campaign targeted repositories including Tiledesk (9 repositories), Black-Iron-Project (8 repositories), and hundreds of others. @tiledesk/tiledesk-server versions 2.18.6-2.18.12 on npm carry the backdoor. C2: 216.126.225.129:8443.
- [Polymarket npm Packages Steal Cryptocurrency Wallet Keys (2026-05-21)](https://safedep.io/blog/malicious-polymarket-npm-crypto-wallet-drainer.md): Nine coordinated npm packages target Polymarket traders with a social-engineered postinstall prompt that exfiltrates raw private keys to a Cloudflare Worker. The attacker published all packages within 30 seconds from a throwaway account.
- [art-template npm Hijack Delivers iOS Browser Exploit Kit (2026-05-20)](https://safedep.io/blog/art-template-npm-supply-chain-compromise.md): art-template versions 4.13.3 through 4.13.7 were compromised after the original author was tricked into transferring ownership through an acquisition fraud. The browser bundle injects scripts that previously delivered the Coruna iOS exploit kit and now redirect visitors to Chinese gambling portals.
- [Malicious durabletask on PyPI: Multi-Cloud Credential Stealer with Worm Capabilities (2026-05-20)](https://safedep.io/blog/malicious-durabletask-pypi-supply-chain-attack.md): Three compromised versions of the Microsoft durabletask Python SDK (1.4.1, 1.4.2, 1.4.3) were published to PyPI, each downloading a stage-2 payload that steals credentials from AWS, Azure, GCP, Kubernetes, HashiCorp Vault, and password managers, then propagates to other hosts via SSM and kubectl exec.
- [Mini Shai-Hulud Strikes Again: 317 npm Packages Compromised (2026-05-19)](https://safedep.io/blog/mini-shai-hulud-strikes-again-314-npm-packages-compromised.md): A compromised npm maintainer account published 637 malicious versions across 317 packages including size-sensor, echarts-for-react, timeago.js, and hundreds of @antv scoped packages, affecting 15M+ monthly downloads.
- [Compromised node-ipc on npm: Credential Stealer via DNS Exfiltration (2026-05-14)](https://safedep.io/blog/malicious-node-ipc-npm-compromise.md): Analysis of compromised node-ipc versions 9.1.6, 9.2.3, and 12.0.1 on npm: a maintainer account takeover injects an 80KB obfuscated credential stealer that targets 100+ sensitive files (SSH keys, cloud credentials, environment variables, AI tool configuration files) and exfiltrates them via DNS tunneling to a C2 server masquerading as Azure infrastructure.
- [Malicious npm Packages Backdoor Claude Code Sessions (2026-05-13)](https://safedep.io/blog/malicious-npm-packages-claude-code-hooks.md): Five typosquatting npm packages ship a hidden ELF binary that fires on install and re-runs via Claude Code's SessionStart hook on every developer session. C2 is 207.90.194.2:443.
- [Cache Poisoning Through pull_request_target: The TanStack Incident (2026-05-13)](https://safedep.io/blog/tanstack-github-actions-cache-poisoning.md): A GitHub user opened a PR against TanStack Router from a fork, poisoned the shared pnpm cache through a pull_request_target workflow, then force-pushed the branch clean. When the release pipeline restored the poisoned cache, the payload executed. Full attack chain analysis with timeline, IOCs, and remediation.
- [Mass Supply Chain Attack Hits TanStack, Mistral AI npm and PyPI Packages (2026-05-12)](https://safedep.io/blog/mass-npm-supply-chain-attack-tanstack-mistral.md): Over 400 compromised npm package versions and at least 2 PyPI packages published in a coordinated supply chain attack targeting TanStack, Mistral AI, UiPath, OpenSearch, guardrails-ai, and dozens of other packages.
- [Endpoint Protection for Developer Machines (2026-05-11)](https://safedep.io/blog/endpoint-protection-developer-security.md): PMG blocks malicious package installs before post-install scripts run. Sync with SafeDep Cloud for fleet-wide visibility across your team's endpoints and CI runners.
- [noon-contracts npm Package: DeFi Supply Chain RAT (2026-05-10)](https://safedep.io/blog/malicious-noon-contracts-npm-defi-rat.md): noon-contracts poses as a Noon Protocol SDK on npm. On install it exfiltrates SSH keys, cryptocurrency wallet private keys, AWS credentials (including live STS/S3/SecretsManager calls), Kubernetes secrets, .env files, shell history, and browser wallet paths to C2 at 82.221.101.203:8443. A full eval-based remote shell polls every 45 seconds. Triple persistence via crontab, macOS LaunchAgent, Linux systemd, and shell RC injection.
- [martinez-polygon-clipping-tony: Trojanized npm Fork Drops Telegram RAT (2026-05-07)](https://safedep.io/blog/malicious-martinez-polygon-clipping-tony-npm-telegram-rat.md): martinez-polygon-clipping-tony is a trojanized fork of the legitimate martinez-polygon-clipping npm package. The postinstall hook downloads a PyInstaller-packed Telegram bot from 172.86.73.132 that provides full remote shell, screenshot capture, file upload/download, and self-destruct capabilities on Windows targets.
- [node-env-resolve: npm Package Installs a Full RAT (2026-05-03)](https://safedep.io/blog/malicious-npm-node-env-resolve-rat.md): node-env-resolve is a malicious npm package that installs a full-featured remote access trojan on developer machines. The RAT streams screens, captures audio, steals browser history, and gives full mouse and keyboard control to a remote operator. The toolkit matches the OtterCookie RAT family linked to North Korea's Contagious Interview campaign.
- [common-tg-service: 502 npm Versions Hijack Telegram (2026-05-01)](https://safedep.io/blog/malicious-common-tg-service-npm-telegram-hijacking-framework.md): common-tg-service ships 502 npm versions of a Telegram account-takeover framework with hardcoded 2FA credentials, IMAP-based code harvesting, and forced session eviction. Its companion package ams-ssk is the server-side runtime.
- [exiouss: Cookie Stealer Bundled in npm Exam Cheat (2026-05-01)](https://safedep.io/blog/malicious-exiouss-npm-exam-cheating-tool.md): exiouss on npm is the latest package from the loltestpad campaign — the same attacker who published the ixpresso-core Windows RAT in April. It bundles a dormant ChatGPT cookie stealer alongside an AI exam cheating tool, targeting students who willingly run it.
- [PyTorch Lightning Compromised: Shai-Hulud Worm Reaches PyPI (2026-04-30)](https://safedep.io/blog/malicious-pytorch-lightning-pypi-compromise.md): PyPI yanked PyTorch Lightning versions 2.6.2 and 2.6.3 after both embedded a two-stage credential-stealing payload. Any import of the library spawns an 11MB obfuscated JavaScript worm identical to the Shai-Hulud payload seen in the April 29 SAP npm campaign.
- [Mini Shai Hulud and SAP Compromise (2026-04-29)](https://safedep.io/blog/mini-shai-hulud-and-sap-compromise.md): Four SAP npm packages published on April 29, 2026 contain a two-stage credential-stealing payload targeting GitHub tokens, AWS keys, and CI/CD pipelines. The packages share SAP-affiliated maintainers, pointing to a publisher account compromise.
- [Malicious redeem-onchain-sdk npm Targets Cryptocurrency Wallets (2026-04-29)](https://safedep.io/blog/redeem-onchain-sdk-polymarket-npm-malware.md): redeem-onchain-sdk impersonates a Polymarket helper SDK and exfiltrates SSH keys, AWS credentials, npm tokens, Docker configuration files, Chrome saved logins, and a month of local git history to an AWS-hosted C2. The payload was shipped dormant for nearly a month, then weaponized with a postinstall hook in four rapid republishes on April 29, 2026.
- [npm-global-util: Credential Theft and Supply Chain Attack (2026-04-29)](https://safedep.io/blog/npm-global-util-malicious-package-analysis.md): npm-global-util is a malicious npm package by maintainer raya4321 that exfiltrates credentials and system recon data via a preinstall hook. Part of a 16-package campaign targeting Apple developer CI/CD environments, with a second-stage that attempts to poison apple-app-store-server-library.
- [Bitwarden CLI Supply Chain Compromise (2026-04-24)](https://safedep.io/blog/bitwarden-cli-supply-chain-compromise.md): A technical writeup of the malicious `@bitwarden/cli@2026.4.0` release linked to the Checkmarx campaign. Covers the poisoned publish path, loader changes, credential theft, GitHub abuse, and responder takeaways.
- [Malicious Pull Requests: A Threat Model (2026-04-22)](https://safedep.io/blog/malicious-pull-requests-threat-model.md): A compact threat model of the malicious pull request as a supply chain attack primitive against GitHub Actions: attacker, goals, assets, controllable surface, and an attack vector taxonomy (V1 through V8).
- [ixpresso-core: Windows RAT Disguised as a WhatsApp Agent (2026-04-16)](https://safedep.io/blog/malicious-ixpresso-core-npm-rat.md): ixpresso-core poses as an AI WhatsApp agent on npm but installs Veltrix, a Windows RAT that steals browser credentials, Discord tokens, and keystrokes via a hardcoded Discord webhook.
- [PMG dependency cooldown: wait on fresh npm versions (2026-04-16)](https://safedep.io/blog/pmg-dependency-cooldown.md): Package Manager Guard (PMG) blocks malicious installs and now supports dependency cooldown, a configurable window that hides brand-new npm versions during resolution so installs prefer older, already-visible releases.
- [forge-jsx npm Package: Purpose-Built Multi-Platform RAT (2026-04-15)](https://safedep.io/blog/malicious-forge-jsx-npm-rat.md): forge-jsx poses as an Autodesk Forge SDK on npm. On install it deploys a system-wide keylogger, recursive .env file scanner, shell history exfiltrator, and a WebSocket-based remote filesystem backdoor to C2 at 204.10.194.247, with persistence via systemd, LaunchAgent, and Task Scheduler.
- [Malicious npm Package js-logger-pack Ships a Multi-Platform WebSocket Stealer (2026-04-15)](https://safedep.io/blog/malicious-js-logger-pack-npm-stealer.md): js-logger-pack spent three weeks on npm evolving from a probe into a full infostealer and then a binary dropper. Early versions installed an SSH backdoor, hijacked Telegram sessions, drained 27 cryptocurrency wallets, and deployed a cross-platform keylogger. After disclosure on April 15, the attacker pivoted to a HuggingFace-hosted binary dropper named MicrosoftSystem64, now at v1.1.26 with 29 total releases.
- [Malicious dom-utils-lite npm SSH Backdoor via Supabase (2026-04-14)](https://safedep.io/blog/malicious-dom-utils-lite-npm-ssh-backdoor.md): dom-utils-lite and centralogger on npm inject attacker SSH keys into ~/.ssh/authorized_keys and exfiltrate server metadata to Supabase-hosted C2 infrastructure, granting persistent remote access.
- [Malicious npm Dependency Confusion Campaign Targets Genoma UI and Others (2026-04-10)](https://safedep.io/blog/malicious-genoma-ui-npm-dependency-confusion-campaign.md): A dependency confusion campaign by npm user victim59 targets at least three organizations through scoped packages @genoma-ui/components, @needl-ai/common, and rrweb-v1. The packages use install hooks to beacon system reconnaissance data to a DigitalOcean C2 server.
- [big.js Typosquat Campaign Implants SSH Backdoors (2026-04-09)](https://safedep.io/blog/malicious-sjs-biginteger-npm-ssh-theft.md): Three waves of big.js typosquats (sjs-biginteger, bjs-biginteger, cjs-biginteger) from throwaway npm accounts implant SSH backdoors and exfiltrate credentials to Cloudflare-disguised C2 infrastructure.
- [@fairwords npm Packages Hit by Credential Worm (2026-04-08)](https://safedep.io/blog/malicious-fairwords-npm-credential-worm.md): Three @fairwords npm packages were compromised with a self-propagating worm that harvests credentials, cryptocurrency wallets, Chrome passwords, and spreads to other packages using stolen npm tokens.
- [Malicious @velora-dex/sdk Delivers Go RAT via npm (2026-04-08)](https://safedep.io/blog/malicious-velora-dex-sdk-npm-compromised-rat.md): Version 9.4.1 of @velora-dex/sdk, a DeFi SDK with ~2,000 weekly downloads, was compromised to deliver a Go-based remote access trojan (minirat) targeting macOS developers.
- [Malicious hermes-px on PyPI Steals AI Conversations (2026-04-06)](https://safedep.io/blog/malicious-hermes-px-pypi-ai-conversation-stealer.md): hermes-px on PyPI steals AI conversations via triple-encrypted exfiltration to Supabase, routing through a hijacked university endpoint while injecting a stolen 245KB system prompt.
- [prt-scan: A 5-Phase GitHub Actions Credential Theft Campaign (2026-04-03)](https://safedep.io/blog/prt-scan-github-actions-exfiltration-campaign.md): A throwaway GitHub account submitted 219+ malicious pull requests in a single day, each carrying a 352-line payload that steals CI secrets, injects workflows, bypasses label gates, and scans /proc for credentials. Five payload variants target GitHub Actions, npm, and Python ecosystems. First public disclosure.
- [Thirty-Six Malicious npm Strapi Packages Deploy Redis RCE, Database Theft, and Persistent C2 (2026-04-03)](https://safedep.io/blog/malicious-npm-strapi-plugin-events-c2-agent.md): A coordinated campaign of thirty-six malicious npm packages published by four sock-puppet accounts (umarbek1233, kekylf12, tikeqemif26, and umar_bektembiev1) targets Strapi CMS deployments with eight distinct payloads evolving from Redis RCE exploitation and Docker container escape through direct PostgreSQL database theft with hardcoded credentials to fileless reverse shells targeting a cryptocurrency payment platform's Jenkins CI pipeline.

- [Compromised npm Package mgc Deploys Multi-Platform RAT (2026-04-03)](https://safedep.io/blog/malicious-npm-mgc-compromised-rat.md): The npm package mgc was compromised via account takeover, with four malicious versions published in rapid succession deploying a full Remote Access Trojan targeting macOS, Windows, and Linux.
- [Malicious npm Package express-session-js Drops Full RAT Payload (2026-04-02)](https://safedep.io/blog/malicious-npm-package-express-session-js.md): A malicious npm package typosquatting express-session fetches and executes a full Remote Access Trojan from a paste service, targeting browser credentials, cryptocurrency wallets, SSH keys, and more.
- [axios Compromised: npm Supply Chain Attack via Dependency Injection (2026-03-31)](https://safedep.io/blog/axios-npm-supply-chain-compromise.md): axios 1.14.1 was published to npm via a compromised maintainer account, injecting a trojanized dependency that executes a multi-platform reverse shell on install. No source code changes in axios itself, just a new entry in package.json.
- [Compromised telnyx on PyPI: WAV Steganography and Credential Theft (2026-03-27)](https://safedep.io/blog/malicious-telnyx-pypi-compromise.md): Analysis of malicious telnyx 4.87.1 and 4.87.2 on PyPI — a package with over 1 million monthly downloads: injected code uses WAV audio steganography to deliver payloads that steal credentials and establish persistence. Attributed to TeamPCP.
- [Malicious litellm 1.82.8: Credential Theft and Persistent Backdoor (2026-03-24)](https://safedep.io/blog/malicious-litellm-1-82-8-analysis.md): Analysis of compromised litellm 1.82.8 on PyPI: a .pth file triggers credential theft, AWS/K8s secret exfiltration, and persistent C2 backdoor on install.
- [sl4x0 Dependency Confusion: 92 Packages Target Fortune 500 (2026-03-24)](https://safedep.io/blog/sl4x0-dependency-confusion-campaign.md): A sustained dependency confusion campaign by the sl4x0 actor likely targets 20+ organizations including Adobe, Ford, Sony, and Coca-Cola with 92+ malicious npm packages exfiltrating developer data via DNS. Updated July 23, 2026 with a Goldman Sachs-targeting wave (gs-uitk-testing-utils, gs-uitk-object-utils) and a second, untargeted install-beacon pair (uploader-frontend, uploader-frontend-legacy).
- [Trivy Supply Chain Compromise: What Happened, What Was Stolen, and How to Respond (2026-03-23)](https://safedep.io/blog/trivy-teampcp-supply-chain-compromise.md): A consolidated technical reference for the TeamPCP supply chain attack against Aqua Security's Trivy scanner. Covers the full attack chain from AI-assisted initial breach through credential theft, GitHub Actions tag poisoning, a self-propagating npm worm, and Kubernetes escape payloads. Includes IOCs, detection queries, and remediation steps.
- [Malicious npm Package react-refresh-update Drops Cross-Platform Trojan on Developer Machines (2026-03-16)](https://safedep.io/blog/malicious-npm-react-refresh-update.md): A malicious npm package impersonating react-refresh, Meta's library with 42 million weekly downloads, was detected by SafeDep. The package injects a two-layer obfuscated dropper into runtime.js that silently fetches and executes a platform-specific second-stage payload from malicanbur[.]pro on require(), targeting Windows, Linux, and macOS.

- [How to Write Time-Based Security Policies in SafeDep vet (2026-03-10)](https://safedep.io/blog/writing-time-based-policies-in-vet-cel.md): Protect against unknown malicious open source packages by enforcing a supply chain cooling-off period using the now() CEL function in SafeDep vet.
- [Threat Modeling the AI-Native SDLC: Supply Chain Security in the Age of Coding Agents (2026-03-09)](https://safedep.io/blog/ai-native-sdlc-supply-chain-threat-model.md): AI agents are rewriting the software development lifecycle. From vibe coding to autonomous CI/CD, every phase now involves an LLM making decisions about your code and dependencies. Here is a threat model for the AI-native SDLC from a supply chain security perspective.
- [Malicious npm Package pino-sdk-v2 Exfiltrates Secrets to Discord (2026-03-06)](https://safedep.io/blog/malicious-npm-package-pino-sdk-v2-env-exfiltration.md): A malicious npm package impersonating the popular pino logger was detected by SafeDep. The package hides obfuscated code inside a legitimate library file to steal environment secrets and send them to a Discord webhook.

- [Gryph: Audit Trail for AI Coding Agents (2026-03-04)](https://safedep.io/blog/gryph-ai-agent-audit-trail.md): AI coding agents operate with broad access to your codebase, credentials, and shell. Gryph logs every action they take to a local SQLite database, making agent behavior visible, queryable, and auditable.
- [Integrate SafeDep MCP in GitHub Agentic Workflow (2026-02-27)](https://safedep.io/blog/safedep-mcp-in-github-agentic-workflow.md): Learn how to integrate SafeDep MCP with GitHub Agentic Workflows to automatically evaluate the security posture of OSS dependencies in your pull requests using AI.
- [Shadow AI Discovery: Find Every AI Tool and SDK in Your Stack (2026-02-27)](https://safedep.io/blog/shadow-ai-discovery-vet.md): AI tools and SDKs are spreading across developer environments faster than security teams can track. vet discovers agents, MCP servers, extensions, and AI SDK usage in code. Open source, local, one CLI.
- [Malicious npm Packages Target Schedaero via Dependency Confusion (2026-02-25)](https://safedep.io/blog/schedaero-dependency-confusion-attack.md): A detailed analysis of a dependency confusion supply chain attack likely targeting Schedaero, a leading aviation software company. We dissect the payload, the exfiltration mechanism, and the indicators of compromise.
- [npm SANDWORM_MODE Attack: Step-by-Step Malware Analysis (2026-02-21)](https://safedep.io/blog/npm-sandworm-mode-supply-chain-attack.md): Step-by-step technical analysis of the SANDWORM_MODE npm supply chain attack. We dissect yarsg and format-defaults malicious packages, decode multi-layer obfuscation, and trace the payload delivery chain.
- [AI Agent Cline v2.3.0 Compromised: From Prompt Injection to Unauthorized npm Publish (2026-02-18)](https://safedep.io/blog/cline-cli-compromised.md): A compromised npm token was used to publish a tampered version of Cline CLI. A prompt injection vulnerability in Cline's AI-powered GitHub Actions workflow may have enabled the credential theft.
- [Why We Built a Hosted MCP Server to Stop Malicious Packages for AI Agents (2026-02-16)](https://safedep.io/blog/why-we-built-a-hosted-mcp-server-for-ai-coding-agents.md): Exposing an MCP server is trivial. Making it useful for AI agents is not. Here's what we learned dogfooding our own tool, and why we built a hosted MCP server backed by real-time open source threat intelligence.

- [End-to-End test with Next.js, Playwright and MSW (2026-02-03)](https://safedep.io/blog/end-to-end-test-nextjs-msw-playwright.md): A practical Next.js 16 App Router E2E setup with Playwright and MSW that keeps server-side fetch deterministic by focusing mocking where it matters, not on server actions.
- [Agent Skills Threat Model (2026-01-23)](https://safedep.io/blog/agent-skills-threat-model.md): Discover critical security threats in Agent Skills - Anthropic's open format for AI agent capabilities. Learn about supply chain attacks, deferred code execution, prompt injection, and multiple attack vectors. Essential reading for developers and security teams implementing AI agents.
- [The State of MCP Registries (2025-12-20)](https://safedep.io/blog/the-state-of-mcp-registries.md): Explore the architecture of the Model Context Protocol (MCP) and the state of its official registry. Learn how to consume server packages programmatically and discover the underlying challenges of data duplication and security in the current meta-registry ecosystem.
- [DarkGPT: Malicious Visual Studio Code Extension Targeting Developers (2025-12-10)](https://safedep.io/blog/dark-gpt-vscode-malicious-extension.md): Malicious extensions are lurking in the Visual Studio Code marketplace. In this case, we discover and analyze DarkGPT, a Visual Studio Code extension that exploits DLL hijacking to load malicious code through a signed Windows executable.
- [Unpacking CVE-2025-55182: React Server Components RCE Exploit Deep Dive and SBOM-Driven Identification (2025-12-04)](https://safedep.io/blog/react-server-nextjs-critical-vulnerability-find-and-fix-with-sbom.md): A critical pre-authenticated remote code execution vulnerability (CVE-2025-55182) was disclosed in React Server Components, affecting Next.js applications using the App Router. Learn about the technical details of this prototype pollution vulnerability, how to identify affected applications using SBOM queries.
- [Shai-Hulud 2.0 npm Supply Chain Attack Technical Analysis (2025-11-24)](https://safedep.io/blog/shai-hulud-second-coming-supply-chain-attack.md): Critical npm supply chain attack compromises zapier-sdk, @asyncapi, posthog, and @postman packages with self-replicating malware. Technical analysis reveals credential harvesting, GitHub Actions exploitation, and worm-like propagation affecting 25,000+ repositories. Includes IOCs, detection methods, and remediation steps.
- [An Opinionated Approach for Frontend Testing for Startups (2025-10-28)](https://safedep.io/blog/frontend-testing-guide.md): How we test our Frontend applications powered by React Query and server components with Vitest.
- [Curious Case of Embedded Executable in a Newly Introduced Transitive Dependency (2025-10-27)](https://safedep.io/blog/curious-case-of-dependency-change-with-embedded-binary-stringish.md): A routine dependency upgrade introduced a suspicious transitive dependency with an embedded executable. While manual analysis confirmed it wasn't malicious, this incident highlights the implicit trust we place in open source code and how attackers exploit the software supply chain through seemingly innocent dependency changes.
- [Malicious npm Packages Impersonating Hyatt Internal Dependencies (2025-10-23)](https://safedep.io/blog/malicious-npm-packages-hyatt-campaign.md): Three malicious npm packages disguised as Hyatt internal dependencies were discovered using install hooks to execute malicious payloads. All packages share identical attack patterns and infrastructure.
- [Contributing to SafeDep Open Source Projects during Hacktoberfest 2025 (2025-10-06)](https://safedep.io/blog/hacktoberfest-safedep-2025.md): Learn how to contribute to SafeDep open source projects during Hacktoberfest 2025 and help secure the open source software supply chain.
- [Ship Code. Not Malware. SafeDep Launches GitHub App for Malicious Package Protection (2025-09-25)](https://safedep.io/blog/ship-code-not-malware-safedep-launches-github-app.md): SafeDep launches a GitHub App for zero-configuration protection against malicious open source packages. Instantly scan pull requests and keep your code repositories safe from supply chain attacks.
- [Shai-Hulud Supply Chain Attack Incident Response (2025-09-22)](https://safedep.io/blog/shai-hulud-supply-chain-attack-response.md): The Shai-Hulud supply chain attack is a major incident targeting
developers through malicious packages in the npm ecosystem. This post
outlines the incident response steps that can be taken to contain and
mitigate the impact of this attack.

- [Diff-based SCA with AI is Broken — Real Examples from Pipfile.lock, yarn.lock, and Cargo.lock (2025-09-19)](https://safedep.io/blog/pitfalls-of-diff-based-sca-scanners.md): Diff-based Software Composition Analysis (SCA) scanners in pull requests are prone to blind spots. By relying only on git diff data,
they miss package context, suffer from nondeterministic rearrangements, and can be trivially bypassed—leaving vulnerabilities undetected.

- [npm Supply Chain Attack Exposes Private Repositories, AWS Credentials and More (2025-09-16)](https://safedep.io/blog/npm-supply-chain-attack-targeting-maintainers.md): npm supply chain attacks continue. This time targeting @ctrl/tinycolor and multiple other packages with credential stealer malware. In this blog, we will analyze the attack and its impact on the npm ecosystem. We will also look at common attack patterns that are being used to target maintainers.
- [npm Supply Chain Attack: Multiple Popular Packages Hijacked (1B+ Weekly Downloads) (2025-09-08)](https://safedep.io/blog/multiple-npm-packages-compromised-billion-downloads.md): Complete analysis of sophisticated cryptocurrency wallet drainer found in 21 npm packages with over one billion weekly downloads. Includes detailed technical breakdown of 76KB malware payload disguised in has-ansi@6.0.1 and multi-stage attack architecture.
- [nx Build System Compromised Targeting Linux and macOS developers (2025-08-27)](https://safedep.io/blog/nx-build-system-compromise.md): The popular npm package `nx` was compromised, targeting Linux and macOS developers. Malicious versions included a postinstall script that stole credentials, exfiltrated sensitive files, and added destructive commands to shell configuration files, causing system shutdowns and data leaks.
- [TensorFlow.js Typosquatting Attack: Malicious Package Targeting AI/ML Developers (2025-08-12)](https://safedep.io/blog/malicious-npm-package-targeting-tensorflow-users.md): A malicious npm package targeting TensorFlow users was discovered on npm. The package uses typosquatting to target the popular `tensorflow` package.
- [Multiple Malicious Python Packages Targeting Bittensor Cryptocurrency Developers (2025-08-12)](https://safedep.io/blog/malicious-python-packages-target-crypto-developers.md): Multiple malicious Python packages targeting cryptocurrency developers and their applications using typosquatting were discovered on PyPI. The packages were used to steal funds by executing a stealthy staking operation.
- [Security Risks in PEP 723 and uv: Inline Metadata Gone Wrong? (2025-08-01)](https://safedep.io/blog/pep-723-inline-metadata-security.md): PEP 723 introduces inline metadata for Python scripts, making tools like `uv` more convenient—but also potentially more dangerous. This post explores security pitfalls when dependencies are declared inside code files.

- [Secure Vibe Coding with AI Agents (2025-07-25)](https://safedep.io/blog/vibe-coding-without-getting-pwned.md): AI coding agents make development faster but can inadvertently introduce security risks by suggesting unvetted packages. Learn how to use vet MCP server for adding security to your vibe coding adventures.
- [eslint-config-prettier Compromised: How npm Package with 30 Million Downloads Spread Malware (2025-07-21)](https://safedep.io/blog/eslint-config-prettier-major-npm-supply-chain-hack.md): A supply chain attack exploiting eslint-config-prettier and other popular npm packages were discovered with major supply chain impact. In this blog, we will explore the details of the hack and the impact it had on the npm ecosystem.
- [SBOM Completeness with Direct & Transitive Dependencies (2025-07-05)](https://safedep.io/blog/sbom-direct-transitive-deps.md): Hidden transitive dependencies create security blind spots. This blog
shows developers and CISOs how SafeDep vet uncovers full Maven dependency
graphs, generating CycloneDX SBOMs and compliance-ready visuals.

- [SBOM and the EU Cyber Resilience Act (CRA) – What Software Vendors Need to Know (2025-06-13)](https://safedep.io/blog/sbom-and-eu-cra-cyber-resilience-act.md): The EU Cyber Resilience Act makes SBOMs mandatory for software products sold in Europe starting December 2027, with fines up to €15 million for non-compliance. Here's what software vendors need to know and how to prepare.
- [Introducing SafeDep Model Context Protocol (MCP) Server to Secure AI Generated Code (2025-06-06)](https://safedep.io/blog/introducing-vet-mcp-server.md): Introducing SafeDep Model Context Protocol (MCP) Server, a new feature in SafeDep vet to secure AI generated code and protect against slopsquatting attacks, vulnerable and malicious packages.
- [License Compliance with SBOM (2025-06-05)](https://safedep.io/blog/license-compliance-with-sbom.md): Although open-source speeds up development, there are risks associated with licensing. This blog examines the ways in which Software Bills of Materials, or SBOMs, facilitate audits, enforce license compliance, and identify infractions early. Discover how to use tools like Vet to incorporate license checks into your DevSecOps pipeline.
- [Introducing Container Image Scanning (2025-06-03)](https://safedep.io/blog/introducing-container-scanning.md): Introducing Container Image Scanning, a new feature in vet to identify vulnerabilities and malicious packages in container images.
- [Software Bill of Materials: Foundation for Trust in Software Supply Chain (2025-06-01)](https://safedep.io/blog/software-bill-of-materials-sbom.md): Modern software rarely ships as a single, hand-crafted binary. Instead, it is assembled from hundreds, sometimes thousands of third-party components that evolve on their own schedule. Knowing exactly what went into an application is now a basic security expectation, and the Software Bill of Materials (SBOM) is how that knowledge is expressed, stored and shared.
- [Catching the Silent Threat: How Dynamic Analysis Revealed a Complex npm Attack Chain (2025-05-19)](https://safedep.io/blog/digging-into-dynamic-malware-analysis-signals.md): Explore how analyzing runtime behaviors using Dynamic Analysis data helps uncover abnormal activities in open source packages. By examining  network connections and unusual binary executions during package installation, we identify potential malicious actors and packages.
- [Introducing Package Manager Guard (PMG) (2025-05-15)](https://safedep.io/blog/introducing-package-manager-guard.md): Introducing Package Manager Guard (PMG), a new tool to protect developers from malicious packages at the time of installation. Seamless integration with popular package managers like npm, pnpm etc.
- [Dynamic Malware Analysis of Open Source Packages at Scale (2025-05-01)](https://safedep.io/blog/dynamic-analysis-oss-package-at-scale.md): Exploring the idea of building a complementary system that can verify and correlate static analysis findings. That's where dynamic analysis comes in, the ability to "run" an open source package in an observed environment and determine its safety status based on real behavior at runtime
- [Malicious npm Package Impersonating Popular Express Cookie Parser (2025-04-23)](https://safedep.io/blog/malicious-npm-package-express-cookie-parser.md): A malicious npm package impersonating the popular Express cookie parser package was discovered by SafeDep Cloud malicious package scanning service.
- [Malicious npm Package Impersonating Java SLF4J (2025-04-21)](https://safedep.io/blog/malicious-npm-package-impersonating-slf4j.md): A malicious npm package impersonating the popular Java logging framework SLF4J is discovered by SafeDep Cloud malicious package scanning service.
- [Announcing DefectDojo Integration (2025-04-14)](https://safedep.io/blog/vet-defect-dojo-integration.md): Introducing DefectDojo Integration allowing vet users to export scan results to DefectDojo. Continue leveraging DefectDojo for your vulnerability management while using vet for identifying vulnerable and malicious open source packages.
- [Analysis of 5000+ Malicious Open Source Packages (2025-04-10)](https://safedep.io/blog/malysis-evaluation-using-datadog-malicious-packages-dataset.md): Analysis of malicious open source packages from Datadog's malicious packages dataset. Each of these packages were found in the wild and confirmed to be malicious. The goal of this analysis is to understand the nature of malicious OSS packages and how they are distributed in the wild.
- [🚀 Introducing GitLab CI/CD Component (2025-03-31)](https://safedep.io/blog/introducing-gitlab-ci-component.md): Introducing GitLab CI/CD Component, available in GiLab CI Catalog for seamless integration of vet in GitLab CI. Protect against vulnerable and malicious packages in your GitLab projects.
- [Agentic Workflows for Malicious Package Analysis (2025-03-28)](https://safedep.io/blog/agentic-workflows-for-malicious-package-analysis.md): Experiments with agentic workflows for malicious package analysis built using Claude Desktop, Model Context Protocol (MCP) server, static code analysis and SafeDep Cloud API tools.
- [Typosquat alert! Malicious npm Package: nyc-config (2025-03-13)](https://safedep.io/blog/nyc-config-malicious-package.md): Possible typosquatting against @istanbuljs/load-nyc-config with ~25M weekly downloads.
- [Introducing vetpkg.dev - Open Source Component Security Dashboard (2025-02-18)](https://safedep.io/blog/introducing-vetpkg-dev.md): Introducing vetpkg.dev - Built using SafeDep API to provide an easy to use visibility of open source component security information.
- [Eliminating SCA Noise using Dependency Usage Evidence (2025-02-07)](https://safedep.io/blog/vet-code-scanning-dependency-usage-evidence.md): SafeDep Code Analysis framework augments vet, our free and open source tool with code context.
- [What is Next Generation Software Composition Analysis? (2025-02-06)](https://safedep.io/blog/what-is-next-gen-sca.md): Software Composition Analysis has been there for a while. But the problems associated with open source vulnerabilities persist. Next-gen SCA is the promised solution. What is it and how does it work?
- [Malicious npm Packages using Burp Collaborator for Dependency Confusion Attack (2025-01-16)](https://safedep.io/blog/burp-collaborator-for-dependency-confusion-attack.md): Multiple npm packages impersonating popular package names were published to the npm registry including by a Snyk researcher apparently targeting internal packages at Cursor AI.
- [npm - The Playground for Malicious Packages (2024-12-11)](https://safedep.io/blog/multiple-npm-malicious-package-impersonating-popular-names.md): Multiple npm packages impersonating popular package names are being used to distribute malware. We take a closer look at the campaign.
- [Malicious Open Source Library Analysis: llm-oracle and its Payload (2024-11-04)](https://safedep.io/blog/malicious-oss-package-analysis-llm-oracle.md): Malware hidden in open source library packages are real. In this article, we analyse the malicious npm package llm-oracle.
- [How a Security Team use Policy as Code for Open Source Security (2024-10-22)](https://safedep.io/blog/vet-policy-as-code-accel-cyber-security-summit-2024.md): This is a talk given at Accel Cyber Security Summit 2024 about securing the open source software supply chain using SafeDep vet. This talk highlights a case study of using policy as code for setting up guardrails
- [SQL Query Interface over SBOM using SafeDep Cloud (2024-10-18)](https://safedep.io/blog/safedep-cloud-preview-sql-query-api.md): This is a '#buildinpublic' update for SafeDep Cloud Development. UI often becomes a bottleneck for developer tools causing friction. We want to overcome it by providing an SQL query interface of SBOM and security metadata.
- [Why Open Source Risks are Larger than SCA Tools (2024-10-03)](https://safedep.io/blog/why-oss-risks-larger-than-sca.md): Open Source Software is critical. However it often comes with inherited risks that are larger than what can be tackled by conventional Software Composition Analysis (SCA) tools.
- [Sneak Peak into SafeDep Cloud Development and SQL Queries (2024-09-30)](https://safedep.io/blog/sneak-peak-into-control-tower-sql-query.md): Software Bill of Material (SBOM) provides an inventory of all software components. However, they are useful only when a flexible query interface is built on top.
- [Safe and Secure Consumption of Open Source Libraries (2024-09-15)](https://safedep.io/blog/safe-and-secure-oss-consumption.md): Open Source software is the foundation of modern software projects. Any software written today consists of 70-90% of open source code in form of libraries and other components.
