In 2016, the Foundation first funded the Community Health Analytics for Open Source Software (CHAOSS) project, led by information scientists Sean Goggins and Matt Germonprez, at the University of Missouri and the University of Nebraska at Omaha (respectively). The CHAOSS project focuses on the development of metrics, software, and practices to improve the transparency of open source community health. With respect to metrics, the CHAOSS project advances open source community health with respect to project evolution, risk, value, and diversity, equity, & inclusion.  With respect to software, the CHAOSS project uses the Augur software developed through this funding to analyze activity logs, and draw comparisons across entire software ecosystems, and identify anomalies within and across projects using machine learning and AI in an explicitly ethical manner. This enables community stakeholders to make judgements about the health of open source communities and projects using a combination of their own knowledge, and the metrics, data analysis, and insights generated through CHAOSS and Augur. Demand for CHAOSS has been substantial, and the program is seeing use by a number of key players in the open source development space, including partners from both academe and practice. The project has also become essential infrastructure for academic studies of open source practices as well as a source of metrics for research software assessment.  This grant provides three years of support for the continued development and expansion of the CHAOSS project.  Grant funds support the ongoing development and maintenance of the CHAOSS metrics and software and the hiring of additional support staff, as well as a set of specific projects in diversity, equity, & inclusion, open science, journalism, and ensuring the safety of critical systems and infrastructure.  Other grant funds will advance the development of funding models to facilitate the long-term, sustainable operation of the project.