admet-prediction

ADMET property prediction for drug candidates. Full pharmacokinetic panel (Caco-2, PPB, clearance, CYP), toxicity (hERG, AMES, DILI), drug-likeness (Lipinski, QED), using RDKit descriptors and TDC models.

Quality

73%

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

Fix and improve this skill with Tessl

tessl review fix ./backend/cli/skills/chemistry/admet-prediction/SKILL.md

Quality

Content

65%

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

The body is highly actionable with executable commands and useful reference tables, but it lacks validation checkpoints for batch operations and fails to link the provided endpoints reference, leaving a detailed file orphaned while inlining overlapping content.

Suggestions

Link the existing reference, e.g. in the Threshold Summary section add "For full endpoint definitions, limitations, and prioritization, see [references/endpoints_guide.md](references/endpoints_guide.md)."

Add validation/verification steps to the batch workflows (e.g. confirm the input CSV parses, check for molecules RDKit failed to read, verify output row count matches input) before reporting results.

Trim the Overview's clinical-trial-failure framing, which restates domain knowledge Claude already has, to improve token efficiency.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with lean script-reference and threshold tables, but the Overview frames domain context Claude largely already knows ("Poor pharmacokinetic and toxicity properties are responsible for roughly 40% of clinical trial failures"), so it could be tightened.	2 / 3
Actionability	Provides concrete, copy-paste-ready commands with real arguments and SMILES examples (e.g. `python scripts/predict_admet.py --input "CC(=O)Oc1ccccc1C(=O)O" --format table`), plus a clear script-reference table mapping purpose to outputs.	3 / 3
Workflow Clarity	Core workflows are numbered with concrete commands, but batch CSV operations lack any validation/verification checkpoints (no input-format check, no "confirm RDKit parsed all molecules" step), so workflow clarity is capped at 2 per the batch-operation guideline.	2 / 3
Progressive Disclosure	A bundle reference file exists (references/endpoints_guide.md) but the body never links to it, while a large inline threshold table duplicates endpoint detail that should live in that separate file — references are present but not signaled, and separable content is inlined.	2 / 3
	Total	9 / 12 Passed

Description

82%

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 specific, domain-rich description that names concrete endpoints and uses the natural vocabulary of medicinal chemists, but it omits an explicit "Use when..." trigger clause, capping completeness. Adding when-to-use guidance would round it out.

Suggestions

Append an explicit trigger clause, e.g. "Use when triaging drug candidates for pharmacokinetic/toxicity liabilities, running ADMET panels, or flagging hERG/AMES/DILI alerts before synthesis."

Lead with the primary action verb (e.g. "Predicts ADMET properties...") to keep the third-person imperative voice consistent.

Dimension	Reasoning	Score
Specificity	Names concrete actions across a full panel — "Full pharmacokinetic panel (Caco-2, PPB, clearance, CYP), toxicity (hERG, AMES, DILI), drug-likeness (Lipinski, QED)" — listing many specific endpoints rather than vague language.	3 / 3
Completeness	It clearly states what the skill does but provides no "Use when..." clause or equivalent explicit trigger guidance, so per the judging guidelines completeness is capped at 2.	2 / 3
Trigger Term Quality	For this specialized domain the terms a medicinal chemist would actually say are present — "ADMET", "drug candidates", "hERG", "Caco-2", "Lipinski", "QED", "clearance", "CYP" — covering the natural vocabulary of the target user.	3 / 3
Distinctiveness Conflict Risk	The ADMET-for-drug-candidates niche with named endpoints (hERG, AMES, DILI, Caco-2) is a clear, distinct trigger set unlikely to fire for unrelated skills.	3 / 3
	Total	11 / 12 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/admet-prediction/SKILL.md
Commit: 52845c3

Reviewed: about 10 hours 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.