If you're filtering compound libraries in drug discovery, this gives you the standard medicinal chemistry toolkit in one place. It wraps Lipinski's Rule of Five, Veber, PAINS filters, and structural alerts from sources like NIBR and Lilly into a clean API with parallelization built in. The functional group detection and complexity metrics are handy for lead optimization when you need to go beyond basic drug-likeness rules. It's opinionated about what makes a good drug candidate, which is exactly what you want when triaging thousands of molecules. The query language for combining filters feels a bit over-engineered, but the core filtering functions are solid and faster than writing your own RDKit patterns.
npx skills add https://github.com/davila7/claude-code-templates --skill medchem