This gives you programmatic access to NIH's Metabolomics Workbench, which hosts over 4,200 metabolomics studies with MS and NMR data. You can search by m/z values with different ion adducts, standardize metabolite names through RefMet, pull study datasets, and filter by experimental parameters like LCMS vs GCMS or sample type. Particularly useful if you're doing biomarker discovery or need to cross-reference metabolites across KEGG, HMDB, and PubChem identifiers. The REST API is straightforward and returns JSON, so you can quickly build queries to find which studies measured specific compounds or identify unknowns from mass spec data. Good coverage of the main use cases for computational metabolomics work.
npx skills add https://github.com/davila7/claude-code-templates --skill metabolomics-workbench-database