rdkit

Cheminformatics toolkit for fine-grained molecular control. SMILES/SDF parsing, descriptors (MW, LogP, TPSA), fingerprints, substructure search, 2D/3D generation, similarity, reactions. For standard workflows with simpler interface, use datamol (wrapper around RDKit). Use rdkit for advanced control, custom sanitization, specialized algorithms.

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/rdkit/SKILL.md

Quality

Content

65%Weight 40%Scale 1-3

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

A highly actionable, code-rich reference with well-signaled bundle files, but it front-loads a very large inline API tour that duplicates the bundled references and lacks structured validation checkpoints in its workflows. Tightening the body into a lean overview and adding explicit validation feedback loops would raise the weaker dimensions.

Suggestions

Slim the body to a concise overview plus quick-start examples, moving the comprehensive API tour into references/api_reference.md (which already exists) to fix both conciseness and progressive_disclosure.

Add explicit numbered workflow sequences with 'validate → fix → retry' checkpoints for batch SDF processing and reaction application, where sanitization failures are common.

De-duplicate the Lipinski rule and Morgan fingerprint examples that currently appear in multiple sections.

Dimension	Reasoning	Score
Conciseness	The body is mostly lean executable code with no concept-padding, but at ~770 lines it carries redundancy (Lipinski computed in section 4 and again in Common Workflows; fingerprint generation repeated across sections and Best Practices) and inlines an API tour that overlaps the bundled references, so it could be tightened.	2 / 3
Actionability	Nearly every section provides fully executable, copy-paste-ready Python with real imports, calls, and parameters (e.g. Chem.MolFromSmiles, rdFingerprintGenerator.GetMorganGenerator, AllChem.EmbedMolecule), matching the anchor for concrete, runnable guidance.	3 / 3
Workflow Clarity	Workflows are present as complete functions with some validation (None checks, DetectChemistryProblems), but batch/destructive operations lack explicit numbered sequencing with validation checkpoints and feedback loops, so per the scoring note workflow_clarity caps at 2.	2 / 3
Progressive Disclosure	The Resources section clearly signals real, one-level-deep reference and script files, but the SKILL.md body itself is a ~770-line inline API tour that overlaps the provided api_reference.md — content that should be split into references remains inline, matching the 'content that should be separate is inline' anchor.	2 / 3
	Total	9 / 12 Passed

Description

82%Weight 40%Scale 1-3

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, well-triggered description with strong distinctiveness that explicitly contrasts rdkit against datamol. Its main weakness is the absence of a clean 'Use when the user mentions...' trigger clause, leaving the 'when' as comparative selection guidance rather than explicit triggers.

Suggestions

Add an explicit 'Use when the user mentions SMILES, SDF, molecular descriptors, fingerprints, similarity, or substructure searching' trigger clause to lift completeness to 3.

Keep the datamol disambiguation but tighten it so the positive triggers for rdkit stand on their own rather than only as a contrast.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'SMILES/SDF parsing, descriptors (MW, LogP, TPSA), fingerprints, substructure search, 2D/3D generation, similarity, reactions' — matching the anchor for enumerating several specific operations rather than vague domain labels.	3 / 3
Completeness	The 'what' is thorough, but the 'when' is framed as comparative skill selection ('For standard workflows... use datamol... Use rdkit for advanced control, custom sanitization, specialized algorithms') rather than explicit user-mention trigger guidance, so per the rubric guideline it caps at 2 instead of 3.	2 / 3
Trigger Term Quality	Covers natural cheminformatics terms a user would actually say — SMILES, SDF, LogP, TPSA, fingerprints, substructure search, similarity, reactions, drug discovery — giving broad coverage rather than jargon-only or generic phrasing.	3 / 3
Distinctiveness Conflict Risk	It actively disambiguates from the sibling datamol skill and carves a clear niche (advanced control, custom sanitization, specialized algorithms), making conflict with other skills unlikely.	3 / 3
	Total	11 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (784 lines); consider splitting into references/ and linking	Warning
metadata_version	'metadata.version' is missing	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	13 / 16 Passed

Repository: synthetic-sciences/openscience
Path: backend/cli/skills/chemistry/rdkit/SKILL.md
Commit: dc00e07

Reviewed: 11 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.