pytdc

Therapeutics Data Commons. AI-ready drug discovery datasets (ADME, toxicity, DTI), benchmarks, scaffold splits, molecular oracles, for therapeutic ML and pharmacological prediction.

Quality

73%

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 ./backend/cli/skills/chemistry/pytdc/SKILL.md

Quality

Content

76%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.

The content is highly actionable with comprehensive executable examples and good progressive-disclosure structure pointing to real bundle files. Its main weaknesses are workflows that omit explicit validation checkpoints for the batch benchmark evaluation and some conceptual prose that could be trimmed for token efficiency.

Suggestions

Add an explicit validation/verification step to the benchmark evaluation workflow (e.g., check prediction shapes match the test set and confirm all 5 seeds produced scores before calling group.evaluate).

Trim the conceptual one-liners per task category and the Overview paragraph to reduce tokens spent restating domain knowledge Claude already has.

Dimension	Reasoning	Score
Conciseness	The body is dense with lean, executable code examples, but the Overview and per-category conceptual one-liners ('Predict pharmacokinetic properties of drug molecules', 'Quantum mechanical properties of molecules') explain concepts Claude largely already knows and could be trimmed.	4 / 5
Actionability	Provides copy-paste-ready, executable examples spanning single_pred, multi_pred, generation, benchmark groups, and oracles, with concrete dataset names and import paths covering the common cases.	5 / 5
Workflow Clarity	Workflows are sequenced (load → split → train → evaluate) but the benchmark 5-seed batch loop and training steps lack explicit validation/verification checkpoints, and training is left as commented-out notional code, leaving checkpoints implicit.	3 / 5
Progressive Disclosure	A clear Resources section enumerates real one-level-deep bundle files (references/.md, scripts/.py) with well-signaled inline pointers, though some inline dataset catalogs duplicate content that also lives in references/datasets.md.	4 / 5
	Total	16 / 20 Passed

Description

70%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.

The description is specific and occupies a distinctive niche with strong domain trigger terms, but it omits an explicit 'Use when…' trigger clause, capping its completeness. Tightening the voice toward concrete actions and adding explicit trigger guidance would raise the score.

Suggestions

Add an explicit trigger clause such as 'Use when working with drug discovery datasets, benchmarking therapeutic ML models, or predicting molecular properties and drug-target interactions.'

Reframe offerings as concrete actions (e.g., 'Load curated ADME/toxicity datasets, run scaffold splits, generate molecules with oracles') instead of a noun list.

Dimension	Reasoning	Score
Specificity	Enumerates concrete offerings — 'AI-ready drug discovery datasets (ADME, toxicity, DTI), benchmarks, scaffold splits, molecular oracles' — but frames them as a noun list rather than concrete action verbs, leaving a minor gap versus the comprehensive anchor.	4 / 5
Completeness	Provides a clear 'what' (datasets, benchmarks, splits, oracles) but lacks an explicit 'Use when…' trigger clause; 'for therapeutic ML and pharmacological prediction' only weakly implies when, so completeness is capped at 3 per the missing-trigger guideline.	3 / 5
Trigger Term Quality	Includes natural domain terms a user would say — 'drug discovery datasets', 'therapeutic ML', 'ADME', 'toxicity', 'DTI' — with good coverage, though a few natural synonyms and file/format cues are missing.	4 / 5
Distinctiveness Conflict Risk	Targets a clearly distinct therapeutics/drug-discovery ML niche with specific identifiers (ADME, DTI, scaffold splits, molecular oracles), creating minimal overlap risk with other skills.	5 / 5
	Total	16 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: synthetic-sciences/openscience
Path: backend/cli/skills/chemistry/pytdc/SKILL.md
Commit: 3a6c3a9

Reviewed: 1 day 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.