Malicious code in ml-data-shared (PyPI)

The ml-data-shared package distributed via PyPI contained malicious code designed to exfiltrate sensitive system information. During package installation, the code would extract basic system data including IP address, username, and all environment variables from the host system.\n\nThe malicious package was identified and attributed to the 2026-07-ml-shared campaign. The attack vector involved distributing the compromised package directly through the PyPI repository, allowing it to reach any user who installed the package during the active distribution period.\n\nThe malicious behavior was documented by the OpenSSF's malicious-packages project, which tracks confirmed malicious packages across major package ecosystems. The package has been cataloged under identifier MAL-2026-11199.\n\nThis incident demonstrates the risk of supply chain compromise through package repositories, where malicious code can be distributed to a broad audience of developers and systems.