| name | chemistry-rdkit |
| description | Molecular analysis and cheminformatics using RDKit — SMILES handling, descriptor calculation, fingerprints, substructure search, similarity, clustering, and drug-likeness screening via domain tools and the execute_chemistry_code tool. |
| states | ["chemistry.molecule_parsed","chemistry.groups_identified","chemistry.descriptors_computed","chemistry.candidates_filtered","chemistry.fingerprints_computed","chemistry.similarity_computed","chemistry.molecules_clustered"] |
Chemistry / RDKit
Use this skill when the user asks about molecules, chemical structures, SMILES,
molecular properties, similarity, substructure matching, or any cheminformatics
task. Code runs in the execute_chemistry_code tool with RDKit auto-imported.
When calling execute_chemistry_code, always set description to a one-line
summary of what the code does (e.g. "Screen 4 molecules for Lipinski drug- likeness"). The description is shown to the user watching the server, so it
should narrate what and why, not low-level syntax.
State Graph Overview
The domain tools form a directed graph of workflows. parse_molecule is the
entry point; downstream tools have prerequisite states that guide workflow
planning.
parse_molecule
→ chemistry.molecule_parsed
│
├── enumerate_functional_groups → chemistry.groups_identified
│
├── compute_descriptors → chemistry.descriptors_computed
│ │
│ └── filter_drug_candidates → chemistry.candidates_filtered
│
└── compute_fingerprints → chemistry.fingerprints_computed
│
├── find_similar_molecules → chemistry.similarity_computed
│
└── cluster_molecules → chemistry.molecules_clustered
