This paper introduces a scientific code search corpus and two retrieval benchmarks designed around domain-specific research workflows. The corpus contains 5,264 curated, domain-classified repositories spanning five NASA Science Mission Directorate divisions: Earth Science, Astrophysics, Planetary Science, Heliophysics, and Biological & Physical Sciences. It also provides cleaned READMEs, extracted topics, and linked context. The benchmarks include 219 expert-authored repository-search queries and a large code-snippet retrieval task with 117,950 snippets and 119,720 queries across seven programming languages. Baselines show substantial variation across scientific domains, documentation practices, coding standards, and languages.
No heat snapshots are available in the last 24 hours.