datamol
Pythonic wrapper around RDKit with simplified interface and sensible defaults. Preferred for standard drug discovery including SMILES parsing, standardization, descriptors, fingerprints, clustering, 3D conformers, parallel processing. Returns native rdkit.Chem.Mol objects. For advanced control or custom parameters, use rdkit directly.
73
66%
Does it follow best practices?
Impact
81%
3.24xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./scientific-skills/datamol/SKILL.mdQuality
Discovery
82%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong, domain-specific description that clearly lists concrete capabilities and uses natural terminology from computational chemistry. Its main weakness is the lack of an explicit 'Use when...' clause, which would help Claude more reliably select this skill. The distinction from direct RDKit usage is a notable strength that reduces conflict risk.
Suggestions
Add an explicit 'Use when...' clause, e.g., 'Use when the user asks about molecular processing, SMILES strings, chemical fingerprints, drug-likeness descriptors, or cheminformatics tasks that don't require custom RDKit parameters.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: SMILES parsing, standardization, descriptors, fingerprints, clustering, 3D conformers, parallel processing. Also specifies return type (native rdkit.Chem.Mol objects) and when to use RDKit directly instead. | 3 / 3 |
Completeness | The 'what' is well covered with specific capabilities and return types. However, there is no explicit 'Use when...' clause or equivalent trigger guidance — the when is only implied by listing the domain and capabilities. Per rubric guidelines, missing explicit trigger guidance caps completeness at 2. | 2 / 3 |
Trigger Term Quality | Includes strong natural keywords a user in this domain would use: 'RDKit', 'SMILES', 'drug discovery', 'fingerprints', 'descriptors', 'conformers', 'clustering', 'standardization'. These are the exact terms a computational chemist would mention. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche — cheminformatics/drug discovery with specific mention of RDKit, SMILES, fingerprints, conformers. Also explicitly distinguishes itself from using RDKit directly ('For advanced control or custom parameters, use rdkit directly'), which reduces conflict risk with a related skill. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
50%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill provides highly actionable, executable code examples covering a comprehensive range of datamol functionality, which is its primary strength. However, it is far too verbose for a SKILL.md overview—it reads more like full library documentation than a concise skill file. Significant content (complete pipelines, ML integration, troubleshooting, SAR analysis) should be moved to the referenced files, and explanatory text that Claude doesn't need should be removed.
Suggestions
Reduce the SKILL.md to ~100-150 lines by moving complete pipeline examples (SAR, virtual screening, ML integration), troubleshooting, and error handling into the referenced files like references/core_api.md or a new references/workflows.md.
Remove explanatory text Claude already knows—e.g., comments like '# Ethanol', explanations of what Lipinski's Rule of Five is, how Counter works, or what ECFP stands for.
Add explicit validation checkpoints in multi-step workflows, e.g., after standardization: 'Verify: check how many molecules were lost: print(f"Kept {len(valid)} of {len(original)}")'.
Consolidate the Best Practices and Error Handling sections into a brief 5-line checklist rather than full code blocks for patterns Claude can generate on its own.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~450+ lines. Explains basic concepts Claude already knows (what SMILES is, what scaffolds are, how to use Counter, how to train a RandomForest). Includes full pipeline examples, ML integration, troubleshooting, and SAR analysis that bloat the document significantly. Much of this could be in reference files or omitted entirely. | 1 / 3 |
Actionability | All code examples are concrete, executable, and copy-paste ready. Functions are shown with real parameters, return values are documented inline, and complete pipelines demonstrate end-to-end usage with specific arguments. | 3 / 3 |
Workflow Clarity | The complete pipeline example has clear numbered steps and the standardization workflow is well-sequenced. However, there are no explicit validation checkpoints—e.g., after standardization there's no step to verify molecule counts or check for unexpected failures before proceeding to descriptor computation. The batch reaction section also lacks validation of outputs. | 2 / 3 |
Progressive Disclosure | References to separate files (references/core_api.md, references/io_module.md, etc.) are well-signaled and one level deep, which is good. However, the SKILL.md itself contains enormous amounts of inline content that should be in those reference files—full SAR analysis, virtual screening pipelines, ML integration, troubleshooting, and error handling sections all belong in references rather than the overview. | 2 / 3 |
Total | 8 / 12 Passed |
Validation
81%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (705 lines); consider splitting into references/ and linking | Warning |
metadata_version | 'metadata.version' is missing | Warning |
Total | 9 / 11 Passed | |
- Repository
- K-Dense-AI/claude-scientific-skills
- Commit
cbcae7b
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.