deepchem

Molecular ML with diverse featurizers and pre-built datasets. Use for property prediction (ADMET, toxicity) with traditional ML or GNNs when you want extensive featurization options and MoleculeNet benchmarks. Best for quick experiments with pre-trained models, diverse molecular representations. For graph-first PyTorch workflows use torchdrug; for benchmark datasets use pytdc.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

80%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, highly actionable skill body that leans on real bundled scripts and one-level-deep references. Its main weakness is the absence of explicit validation/feedback-loop checkpoints in the batch training workflows, which caps workflow clarity.

Suggestions

Add explicit validation checkpoints to the training workflows (e.g. 'after training, check metric X; if below threshold Y, increase epochs or switch architecture') to turn the 'start simple then scale' ladder into a true feedback loop.

Tighten the Overview/version-note paragraph and the eight-item capability enumeration, since both duplicate detail already in references/core_capabilities.md.

Reconcile the workflow counts: the body says 'Three end-to-end workflows' for typical_workflows.md but 'Eight detailed end-to-end workflows' for workflows.md — make the counts and file roles unambiguous so the reader knows which to open.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with executable snippets and tight best-practice patterns, but the Overview/version note and some enumerated capability lists restate things the references already cover; a few sentences could be trimmed.	4 / 5
Actionability	Provides copy-paste-ready CLI invocations for all three bundled scripts with real flags, plus executable Python patterns for splitting, normalization, balancing, and disk datasets covering the common cases.	5 / 5
Workflow Clarity	Multi-step workflows are implied via script examples and a 'start simple then scale' ladder, but there are no explicit validation checkpoints or validate->fix->retry feedback loops, and the destructive/batch cap applies since these are batch training jobs without validation steps.	3 / 5
Progressive Disclosure	Clear overview body with well-signaled one-level-deep references (core_capabilities.md, typical_workflows.md, api_reference.md, workflows.md) that all exist as real files, plus a bundled scripts/ directory; navigation is easy and nothing is nested deeper than one level.	5 / 5
	Total	17 / 20 Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, third-person description that concretely states capabilities and gives explicit, natural trigger conditions plus helpful boundary routing to competing skills. Only minor trigger-term synonyms are missing.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions (featurizers, pre-built datasets, ADMET/toxicity property prediction with traditional ML or GNNs, MoleculeNet benchmarks, pre-trained models) with comprehensive coverage of the library's core surface.	5 / 5
Completeness	Explicitly answers both 'what' (molecular ML with diverse featurizers and pre-built datasets) and 'when' ('Use for property prediction (ADMET, toxicity)... when you want extensive featurization options'), with concrete trigger phrases.	5 / 5
Trigger Term Quality	Strong natural terms ('property prediction', 'ADMET', 'toxicity', 'MoleculeNet benchmarks') but a few common variations users might say (e.g. 'drug discovery', 'QSAR', 'solubility') are absent, keeping it just below comprehensive.	4 / 5
Distinctiveness Conflict Risk	Clear niche (molecular ML / featurization-heavy chemistry) with explicit routing to torchdrug and pytdc for adjacent cases, minimizing conflict risk with related skills.	5 / 5
	Total	19 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: K-Dense-AI/scientific-agent-skills
Path: skills/deepchem/SKILL.md
Commit: 49c6e97

Reviewed: 9 days ago

Table of Contents

Discovery Implementation Validation

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.