CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

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

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./scientific-skills/deepchem/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%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 highly actionable skill body: executable code everywhere, clearly sequenced workflows with evaluation checkpoints, and clean one-level-deep references to real bundle files. Its main cost is redundancy — scaffold splitting and installation instructions each appear multiple times, and the ~580-line body could be noticeably tightened by consolidating repeats and pushing catalog material into the existing reference files.

Suggestions

Consolidate the three separate scaffold-splitting discussions ('Data Splitting', 'Pattern 1', and 'Issue 1') into a single authoritative section with a one-line cross-reference, and merge the duplicate installation instructions from 'Issue 4: Import Errors' and 'Installation Notes'.

Trim the Overview paragraph, which restates the frontmatter description without adding new guidance for the model.

Move catalog-style content — the full model-selection table, featurizer decision tree, and Common Pitfalls catalog — into references/api_reference.md or a dedicated reference file, leaving SKILL.md as a lean overview with pointers.

DimensionReasoningScore

Conciseness

The body is mostly efficient, actionable code rather than padded prose, but it repeats itself: scaffold splitting is taught in three separate places ('Data Splitting', Pattern 1, Issue 1), installation instructions appear in both 'Issue 4: Import Errors' and 'Installation Notes', and the Overview paragraph restates the frontmatter description — fitting the 'mostly efficient but could be tightened' anchor.

3 / 5

Actionability

Nearly every section contains complete, copy-paste-ready code (CSVLoader/SDFLoader usage, featurizer construction, GCNModel/MultitaskRegressor/SklearnModel training, evaluate/predict calls) plus concrete CLI invocations for all three bundled scripts, covering the common molecular ML cases — matching the 'fully executable; copy-paste ready' anchor.

5 / 5

Workflow Clarity

Three end-to-end workflows (A, B, C) are clearly sequenced with numbered steps 1–5, each ending in an explicit evaluation checkpoint, and the Common Pitfalls section provides problem→solution recovery guidance. It stops short of the level-5 anchor because there are no explicit validate-before-proceeding feedback loops within the workflows themselves.

4 / 5

Progressive Disclosure

References are well-signaled and one level deep, with explicit 'When to reference' guidance for references/api_reference.md and references/workflows.md, and all referenced bundle files (references and scripts) actually exist. It is not a 5 because a fair amount of catalog-style content (featurizer decision tree, model selection table, full pitfalls catalog) is inlined in SKILL.md when it could live in the reference files.

4 / 5

Total

16

/

20

Passed

Description

78%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 description with explicit scope, good natural trigger terms, and notably excellent distinctiveness via explicit routing to competing skills (torchdrug, pytdc). Its main weakness is that capabilities are expressed as preference-framed noun phrases rather than concrete, verb-driven actions with task-based triggers, keeping the mid-level dimensions at 4.

DimensionReasoningScore

Specificity

The description names the domain and several concrete capabilities ('property prediction (ADMET, toxicity)', 'MoleculeNet benchmarks', 'pre-trained models', 'diverse featurizers'), but capabilities are stated as noun phrases rather than a comprehensive list of explicit actions, matching the 'several specific actions; minor gaps' anchor rather than the level-5 comprehensive anchor.

4 / 5

Completeness

Both 'what' ('Molecular ML with diverse featurizers and pre-built datasets') and 'when' ('Use for property prediction... when you want extensive featurization options... Best for quick experiments') are present, but the 'when' clause is framed around user preferences rather than concrete task triggers, so it fits the anchor where 'when' could be more explicit rather than the level-5 fully concrete anchor.

4 / 5

Trigger Term Quality

It includes good natural keywords users would actually say ('ADMET', 'toxicity', 'GNNs', 'MoleculeNet', 'pre-trained models'), but misses common variations like 'SMILES', 'solubility', 'drug discovery', or 'molecular property prediction', fitting the 'good keyword coverage; a few natural terms missing' anchor.

4 / 5

Distinctiveness Conflict Risk

It occupies a clear niche and explicitly disambiguates from adjacent skills ('For graph-first PyTorch workflows use torchdrug; for benchmark datasets use pytdc'), giving minimal conflict risk — a direct match for the 'clear niche with distinct triggers' anchor.

5 / 5

Total

17

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (597 lines); consider splitting into references/ and linking

Warning

metadata_version

'metadata.version' is missing

Warning

Total

14

/

16

Passed

Repository

googolme/run0204

Path

scientific-skills/deepchem/SKILL.md

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.