seoklab/nurikit

NuriKit is a serious, well-engineered chem/bioinformatics project. The C++ core, Python bindings, CMake/scikit-build packaging, PyPI/conda support, docs, CI, CodeQL, fuzzing, benchmarks, sanitizer use, and broad C++/Python test coverage all signal unusually mature infrastructure for a research-adjacent toolkit. The molecule, graph, format, geometry, descriptor, and alignment modules suggest a real platform rather than a narrow prototype.

My main critique is public positioning. The README is very short while making a huge claim about succeeding RDKit/OpenBabel. That ambition is compelling, but outside users need clearer examples, a maturity map, supported platforms, and guidance on what is stable today. The docs expose APIs well, but more cookbook workflows would improve adoption. I’d also consider pinning/vendoring the CMake bootstrap dependency instead of downloading “latest.” Overall, the engineering foundation is impressive; the front-door documentation just needs to catch up.