Amazon released a collection of 38 open-source agent skills spanning 11 healthcare and life sciences domains to improve AI reasoning in clinical and research workflows.

The skills encode domain decision procedures and frameworks in structured markdown format that agents can consume at inference time, addressing gaps where foundation models misapply healthcare guidelines.

Agents equipped with these skills demonstrated 70–86 percent win rates in head-to-head comparisons against agents without skills, with the strongest improvements in critical thinking tasks.