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

76%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/deepchem/SKILL.md

Quality

Content

65%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-organized skill body that leans on real bundled references and executable examples for strong actionability and progressive disclosure. Its weaknesses are mild verbosity in overview/version-note prose and the absence of explicit validation checkpoints in the inline batch workflows.

Suggestions

Tighten the Overview paragraph to drop marketing phrasing and remove the Version Note duplication of the frontmatter compatibility line.

Add an explicit validate-fix-retry checkpoint (e.g., confirm scaffold split sizes or metric thresholds before scaling up) to the Start Simple, Then Scale workflow to raise workflow clarity above 3.

Give partial code snippets like the NormalizationTransformer pattern enough context (imports, full flow) to be copy-paste runnable.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with code blocks, lists, and commands, but the marketing-flavored Overview paragraph and the Version Note that duplicates the frontmatter compatibility line could be tightened.	3 / 5
Actionability	Provides executable bash commands for the bundled scripts, real Python GOOD/BAD snippets, and concrete install commands, with only minor gaps where pattern snippets lack surrounding context.	4 / 5
Workflow Clarity	Sequenced guidance exists (Start Simple, Then Scale) but the body lacks explicit validation/feedback checkpoints for batch training and data-splitting workflows, capping it per the batch-operation rule.	3 / 5
Progressive Disclosure	Clear overview with well-signaled, one-level-deep references (core_capabilities.md, typical_workflows.md, api_reference.md, workflows.md) all of which exist as real files, plus a separate scripts/ directory; easy to navigate.	5 / 5
	Total	15 / 20 Passed

Description

87%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, well-targeted description that pairs concrete capabilities with clear usage triggers and unusually good boundary guidance against adjacent skills. Only minor expansion of synonyms and actions would push it higher.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions (featurizers, pre-built datasets, property prediction of ADMET/toxicity, traditional ML or GNNs, pre-trained models) but leaves minor gaps (data loading, splitting) keeping it below comprehensive.	4 / 5
Completeness	Explicitly answers 'what' (molecular ML with diverse featurizers and pre-built datasets) and 'when' (Use for property prediction when you want featurization options and MoleculeNet benchmarks) with concrete trigger phrases.	5 / 5
Trigger Term Quality	Strong domain keywords (property prediction, ADMET, toxicity, MoleculeNet benchmarks, GNNs, pre-trained models) a cheminformatics user would naturally say, though a few natural synonyms (drug discovery, solubility) are absent.	4 / 5
Distinctiveness Conflict Risk	Clear niche (DeepChem-style molecular ML) with explicit boundary guidance steering graph-first PyTorch to torchdrug and benchmark datasets to pytdc, minimizing wrong-skill triggers.	5 / 5
	Total	18 / 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: b2a92ba

Reviewed: 3 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.