New npm Supply-Chain Worm "ChainDrop" Automates Package Infections at Scale A recently discovered npm supply-chain worm, dubbed ChainDrop, has compromised over 400 npm packages, including widely used projects like keyv and cacheable-request. The malware operates as an automated infection system, leveraging stolen developer credentials to propagate malicious code across the software supply chain. ### How ChainDrop Works Once executed on a developer’s workstation or CI/CD pipeline, ChainDrop scans for sensitive credentials, including: - npm and GitHub publishing tokens - Cloud credentials, SSH keys, and Docker/Kubernetes configurations - Terraform state files and environment variables Using stolen npm credentials, the worm identifies all packages the compromised account can publish, then rebuilds, injects malicious code, increments version numbers, and republishes them all while maintaining the original functionality to avoid detection. ### Infection Mechanism & Evasion Tactics ChainDrop modifies a package’s package.json file to include a preinstall hook that executes setup.mjs, a dropper script. The malware checks for the Bun JavaScript runtime, downloading it if absent, and uses it to run an obfuscated payload (mathinit.js). While Bun itself is not compromised, it is exploited as a portable execution environment. Key evasion techniques include: - Locale-based targeting: Exits on Russian-language systems, suggesting intentional geographic filtering. - Memory scraping: A Python-based helper extracts GitHub Actions runner secrets and OpenID Connect tokens from live processes, capturing credentials that may never be written to disk. - Persistence via developer tools: Creates malicious VS Code task files and Claude AI session hooks to re-execute the payload when projects are opened. Researchers also identified dormant persistence mechanisms for macOS (LaunchAgent) and Linux (systemd user service), though these were not actively deployed in observed attacks. ### Impact & Scope ChainDrop’s automated propagation poses a severe risk to: - Developers (compromised workstations) - CI/CD pipelines (injected build processes) - Cloud environments (exposed credentials) - Downstream users (infected dependencies) The worm’s ability to republish legitimate packages with hidden malicious updates makes detection challenging, as infected packages continue to function normally. ### Indicators of Compromise (IOCs) Security researchers have shared the following hashes for detection: - Payloads: - 9fc2570b7cef51c1b8df116d144d11ff4096357be7d2c4c6367cfc2509cf1bcc (MathSymbol.js/mathinit.js) - 54dc7ea54a1317cca0e890a2770630cf7fa6c97813e0cb9d2caa93012b350668 (setup.mjs – first variant) - fd3ca4007b225fdf8de7af4345a19179d5efa8c4bb9205f88cda806e5684b1eb (setup.mjs – second variant) - b27b82afa5f15512f3856e549fb83d873fd0049759a4b62ce64c8d7d4dc2c678

The expiration of the Cybersecurity Information Sharing Act (CISA 2015) has left ReversingLabs—a cybersecurity firm heavily reliant on the law’s Automated Indicator Sharing Program (AIS)—vulnerable to legal liabilities when exchanging cyber threat intelligence. The lapse, driven by Congressional inaction during a government funding standoff, eliminates critical legal protections that previously encouraged voluntary threat data sharing among companies. Without these safeguards, ReversingLabs faces operational disruptions in maintaining its threat repositories, increasing exposure to software supply chain vulnerabilities and AI-enabled attacks.The company’s Chief Trust Officer, Saša Zdjelar, warned that the lapse creates a ‘chilling effect’ on threat intelligence collaboration, forcing firms to withhold data due to litigation risks. This weakens collective cyber defenses, granting adversaries a strategic advantage. Zdjelar also highlighted risks to AI security development, as legal uncertainty may restrict data-sharing essential for training AI-powered defense tools. The broader impact includes escalating costs of data breaches—already the highest globally in the U.S.—and potential long-term erosion of trust in cybersecurity partnerships, leaving critical infrastructure and enterprises more susceptible to sophisticated cyber threats.