Gives you programmatic access to PubChem's 110M+ compound database through PUG-REST API and PubChemPy. You can search by name, CID, SMILES, or molecular formula, retrieve properties like molecular weight and LogP, and run similarity or substructure searches to find related compounds. The bioactivity endpoints let you pull screening assay data, which is useful for early stage drug discovery work. The skill includes helper scripts for batch processing and common queries. Response times can lag on similarity searches since they're asynchronous on PubChem's end, but PubChemPy handles the polling automatically. Solid option if you're doing cheminformatics work and need more than what RDKit gives you locally.
npx skills add https://github.com/davila7/claude-code-templates --skill pubchem-database