Socket disclosed a fresh wave of supply-chain attacks targeting Ruby gems and Go modules: more than 60 typosquatted packages were uploaded to RubyGems and the Go module registry, designed to look like legitimate dependencies developers might pull into a CI pipeline. Once installed, the packages exfiltrate environment variables (which typically include AWS keys, GitHub tokens, and database credentials in CI environments) to attacker-controlled servers. The targeting is deliberate: typosquats picked names close to popular gems and Go libraries. This is the same operational pattern as the SAP npm compromise covered Wednesday, but targeting Ruby and Go ecosystems.
Update on the Mini Shai-Hulud campaign covered April 30: The same supply-chain worm that hit four SAP npm packages on Wednesday spread to two more major packages on Thursday. PyTorch Lightning, an AI training framework with 31,100 GitHub stars and hundreds of thousands of daily downloads, had malicious versions 2.6.2 and 2.6.3 published on PyPI for 42 minutes before being quarantined. Intercom-client, the official Node.js SDK for Intercom (361,510 weekly downloads), was compromised at 14:41 UTC. Intercom traced its compromise to pyannote-audio pulling Lightning as a dependency - showing the worm propagating through stolen credentials from the SAP victims.
Attackers compromised four official SAP npm packages on Wednesday and replaced them with versions that quietly steal developer credentials when installed. The packages - mbt, @cap-js/sqlite, @cap-js/postgres, and @cap-js/db-service - are SAP's open-source tools for cloud application development. Anyone who ran 'npm install' between 09:55 and 12:14 UTC on April 29 had their machine grab GitHub tokens, npm credentials, and AWS, Azure, and GCP secrets, then dump them into public GitHub repositories on the victim's own account. The same attackers (TeamPCP) hit Trivy, Checkmarx, and Bitwarden earlier this year. The malware skips Russian-language systems entirely.
LiteLLM, the popular open-source gateway used to centralize API access for OpenAI, Anthropic, and other AI providers, has a critical pre-authentication SQL injection bug that attackers started exploiting just 36 hours after the security advisory went public. The flaw lets anyone who can reach the proxy port read all the API keys stored inside - including master keys, virtual keys, and provider credentials. The bug was in the bearer-token check: the token was concatenated into a SQL query instead of passed as a parameter. Sysdig saw the first attack at 04:24 UTC on April 26, hitting three tables that hold the most valuable secrets.
At the Asness Summit in Nashville on April 24, former NSA director Tim Haugh and Mandiant founder Kevin Mandia argued Iran's current cyber posture more closely resembles a criminal actor than a sophisticated APT - reliant on dark-web-purchased credentials, basic security gaps, and information operations to amplify modest intrusions. They cited the March 11 Stryker attack as the template: no malware, no zero-day, just legitimate credentials used to abuse MDM and delete data the attacker had permission to delete. Mandia's CISO advice: assume valid credentials for your staff are already on sale and build detection around their misuse.
Attackers compromised Nextend's update infrastructure and pushed a fully weaponized version of Smart Slider 3 Pro (3.5.1.35) through the official WordPress and Joomla update channel on April 7. Sites with auto-updates enabled received a multi-layered remote access toolkit disguised as a legitimate plugin update. The malicious version was live for approximately six hours before detection. Patchstack's analysis found: unauthenticated remote command execution via crafted HTTP headers, a second authenticated backdoor with PHP eval and OS command execution, a hidden administrator account (prefixed wpsvc_) invisible in the admin interface, persistent backdoors planted in the active theme's functions.php and wp-config.php, and automated credential theft sent to an external server. Traditional defenses like firewalls, nonce verification, and role-based access controls are irrelevant here because the malicious code arrived through the trusted update channel. Affected sites should be considered fully compromised.
Attackers compromised a backend API on CPUID's website and replaced the official download links for CPU-Z and HWMonitor with trojanized versions containing the STX RAT. The attack lasted approximately six hours between April 9-10, timed to when the lead developer was on holiday. The malicious packages used DLL sideloading - legitimate CPUID executables (still properly signed) were bundled alongside a malicious CRYPTBASE.dll that masquerades as a standard Windows library. When users launched HWMonitor or CPU-Z, the malicious DLL loaded and deployed the RAT entirely in memory, with four independent persistence paths. The primary goal was browser credential theft, specifically targeting Chrome's IElevation COM interface to dump and decrypt saved passwords. The same threat group previously compromised FileZilla downloads in early March 2026. CPUID's signed original files were not tampered with - this was an infrastructure attack redirecting download links to attacker-controlled Cloudflare R2 storage.
Cisco Talos uncovered a large-scale automated campaign by threat cluster UAT-10608 that exploits React2Shell - a CVSS 10.0 pre-auth RCE flaw in React Server Components used by Next.js. One crafted HTTP request is all it takes to get code execution, no credentials needed. The attackers scan with Shodan and Censys, breach Next.js apps, then deploy the NEXUS Listener framework to harvest database credentials, SSH keys, AWS tokens, Stripe API keys, Kubernetes secrets, and GitHub tokens at scale. At least 766 hosts across multiple cloud providers were compromised within 24 hours.
Researchers at Censys discovered a previously undocumented Russian-origin toolkit called CTRL, distributed through Windows shortcut files disguised as private key folders. Once a victim double-clicks the LNK file, a multi-stage chain deploys credential harvesting through a fake Windows Hello PIN prompt, a keylogger, RDP session hijacking, and reverse proxy tunneling. All stolen data exits through the RDP tunnel, leaving minimal forensic traces compared to traditional command-and-control patterns.
TrustStrike Labs