diffdock

Predict small-molecule binding poses with DiffDock-L (Corso et al. 2023/2024, github.com/gcorso/DiffDock) — blind diffusion docking that places a ligand into a protein pocket without a predefined search box and ranks the samples with a learned confidence model. Reach for this skill to dock a SMILES or SDF against a PDB, to generate ranked 3D poses for a small fragment library, or to get a starting pose for downstream rescoring. DiffDock predicts geometry, not affinity.

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

93%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 a tight, expert-level runbook dense with executable commands and hard-won failure-mode guidance, correctly splitting bulk batch/sequence detail into one real reference file. The only gap is the absence of an explicit post-run validation step in the main workflow.

Dimension	Reasoning	Score
Conciseness	Lean body that assumes Claude's domain competence and earns every token with non-obvious gotchas (YAML-overwrites-CLI, silent ~11-min precompute, --ligand prefix-match accident); no padding explaining what docking or diffusion models are.	5 / 5
Actionability	Fully executable inference command with real flags, a concrete four-column CSV schema, and an error table giving specific remediation (sed the setrlimit constant, set modal.memory: 65536); copy-paste ready and covers the common single-complex, batch, and failure cases.	5 / 5
Workflow Clarity	Single-complex run is clearly sequenced and the batch/sequence-only routes are pushed to a reference; the gotchas and error table serve as implicit validation checkpoints, but there is no explicit 'verify output then proceed' loop after a run.	4 / 5
Progressive Disclosure	Concise overview body with a single, clearly signaled one-level-deep reference (references/workflows.md) that is a real, appropriately scoped file covering batch and sequence-only workflows; easy to navigate.	5 / 5
	Total	19 / 20 Passed

Description

100%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, trigger-rich, and explicitly answers both what the skill does and when to reach for it, while cleanly distinguishing it from affinity predictors and co-folding skills. Voice is third person throughout, with no fluff or over-claims.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'places a ligand into a protein pocket', 'ranks the samples with a learned confidence model', 'generate ranked 3D poses', 'get a starting pose for downstream rescoring' — with comprehensive coverage of the skill's capabilities.	5 / 5
Completeness	Explicitly answers both what ('Predict small-molecule binding poses... blind diffusion docking... ranks the samples with a learned confidence model') and when ('Reach for this skill to dock a SMILES or SDF against a PDB, to generate ranked 3D poses...'), with concrete trigger phrases.	5 / 5
Trigger Term Quality	Covers natural user phrases ('dock a SMILES or SDF against a PDB', 'fragment library', 'rescoring') plus synonyms (poses, docking) and file extensions (SDF, PDB), matching the comprehensive-coverage anchor.	5 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (DiffDock-L blind diffusion docking), states it 'predicts geometry, not affinity', and routes protein–protein/nucleic-acid co-folding to boltz/chai1, minimizing overlap with sibling skills.	5 / 5
	Total	20 / 20 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
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	13 / 16 Passed

Repository: UnicomAI/wanwu
Path: configs/microservice/bff-service/configs/agent-skills/claude-science/diffdock/SKILL.md
Commit: 99df6e4

Reviewed: 6 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.