molfeat

Molecular featurization for ML (100+ featurizers). ECFP, MACCS, descriptors, pretrained models (ChemBERTa), convert SMILES to features, for QSAR and molecular ML.

1.41x

Quality

70%

Does it follow best practices?

Impact

78%

1.41x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./scientific-skills/molfeat/SKILL.md

Evaluation results

88%

68%

Aqueous Solubility Prediction Pipeline

QSAR pipeline with MoleculeTransformer and sklearn integration

Criteria

Without context

With context

FPCalculator import

100%

MoleculeTransformer usage

100%

Parallel processing

100%

sklearn Pipeline

ECFP fingerprint

100%

ECFP radius param

100%

Config saved to file

100%

Config file produced

100%

Config reloaded

100%

Performance metric reported

100%

72%

Molecular Representation Benchmarking for Scaffold Hopping

Multi-featurizer comparison with FeatConcat and ModelStore discovery

Criteria

Without context

With context

FeatConcat used

100%

At least 2 featurizers in concat

100%

Non-fingerprint featurizer included

100%

ModelStore discovery

100%

n_jobs=-1 parallel

At least 3 featurizer types compared

100%

ignore_errors enabled

Feature dimensions reported

100%

Comparison results saved

100%

MoleculeTransformer wraps FeatConcat

76%

Robust Compound Library Featurization for Production Deployment

Error handling, datamol preprocessing, and config reproducibility

Criteria

Without context

With context

ignore_errors enabled

33%

50%

verbose enabled

datamol imported

100%

datamol standardization

100%

datamol salt removal

100%

Config saved to YAML/JSON

100%

Config reloaded

100%

None/invalid filtering

100%

molfeat version logged

100%

n_jobs=-1 parallel

Repository: K-Dense-AI/claude-scientific-skills
Commit: b58ad7e

Evaluated: 27 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Aqueous Solubility Prediction Pipeline Molecular Representation Benchmarking for Scaffold Hopping Robust Compound Library Featurization for Production Deployment

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.