diffdock

Predicts protein-small-molecule binding poses with DiffDock and DiffDock-L from PDB or sequence plus SMILES/SDF/MOL2. Covers batch docking, pose triage, confidence interpretation, and validation. Use for molecular docking and virtual-screening pose generation, not binding-affinity prediction.

Quality

96%

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

Quality

Content

100%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 body is an exemplary procedural skill document: dense with upstream-verified gotchas (YAML-over-CLI override, ESM2 1022-residue truncation, fresh output directories), fully executable commands, validation feedback loops for batch operations, and a well-organized bundle with one-level-deep references. No dimension shows more than minor room for improvement.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence — it mentions ESM2 truncation, SO(2)/SO(3) tables, and argparse abbreviation without explaining basics, and confines time-sensitive version/date information to the dedicated 'Verified scope' section. Every sentence carries a DiffDock-specific fact, caution, or command.	5 / 5
Actionability	All guidance is copy-paste executable: environment setup (git clone/conda/docker), a single-pair inference command with a real aspirin SMILES, batch CSV creation/validation, config copying for sampling changes, and analysis commands — each with the required working directory and input conventions stated.	5 / 5
Workflow Clarity	A clear sequence (setup → checker → prepare traceable inputs → single pair → batch → sampling changes → inspection) with explicit validation checkpoints at every stage (setup_check.py, CSV --validate, analyze_results.py, manifest/pose-count matching) and error-recovery guidance (troubleshooting section, partial-run checkpoint inspection, stale-run contamination detection).	5 / 5
Progressive Disclosure	The body is a concise overview holding only essential facts, with clearly signaled one-level-deep references (parameters_reference.md, workflows_examples.md, confidence_and_limitations.md) linked at their point of use; all referenced scripts, assets, and reference files exist in the bundle, so navigation is complete and shallow.	5 / 5
	Total	20 / 20 Passed

Description

92%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 a strong, third-person statement with concrete capabilities, explicit use-when triggers, an explicit out-of-scope boundary, and named input formats. Its only weakness is slightly incomplete natural-language synonym coverage (e.g., 'dock', 'protein-ligand') in the trigger phrasing.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Predicts protein-small-molecule binding poses', 'batch docking, pose triage, confidence interpretation, and validation' — plus concrete input formats (PDB, sequence, SMILES/SDF/MOL2), giving comprehensive coverage with no significant gaps.	5 / 5
Completeness	Explicitly answers both what ('Predicts protein-small-molecule binding poses... Covers batch docking, pose triage, confidence interpretation, and validation') and when ('Use for molecular docking and virtual-screening pose generation') with concrete triggers, plus a negative boundary ('not binding-affinity prediction').	5 / 5
Trigger Term Quality	Includes natural trigger phrases such as 'molecular docking', 'virtual-screening pose generation', and format keywords (SMILES/SDF/MOL2), but omits common variations users would say such as 'dock this ligand' or 'protein-ligand docking'. It clearly exceeds the 'some relevant keywords' level but falls just short of comprehensive synonym coverage.	4 / 5
Distinctiveness Conflict Risk	A clear niche (DiffDock/DiffDock-L pose generation) with distinct, specific triggers; the explicit exclusion of binding-affinity prediction further separates it from neighboring scoring/affinity skills, so conflict risk is minimal.	5 / 5
	Total	19 / 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/diffdock/SKILL.md
Commit: 1549884

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.