Skills diffdock
๐Ÿ“ฆ

diffdock

Content revision r1 Low Risk โš™๏ธ External commands๐ŸŒ Network access๐Ÿ“ Filesystem accessโšก Contains scripts

Run DiffDock Molecular Docking

Protein-ligand docking requires careful setup, parameter choices, and result interpretation. This skill guides DiffDock workflows for poses, confidence scores, and screening.

Supports: Claude Codex Code(CC)
๐Ÿฅˆ 81 Silver

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request
Review the Skillstore skill "diffdock" from https://skillstore.io/skills/k-dense-ai-diffdock.md and its manifest at https://skillstore.io/api/skills/k-dense-ai-diffdock/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Test it

Using "diffdock". Dock aspirin against a prepared protein target and explain the expected outputs.

Expected outcome:

The response lists required inputs, recommends a single-complex DiffDock run, and explains ranked pose files plus confidence scores.

Using "diffdock". I have 200 ligands for one target protein. Plan a screening workflow.

Expected outcome:

  • Create a batch input table with one row per ligand.
  • Validate protein paths, ligand descriptions, and required columns.
  • Precompute protein embeddings for repeated target use.
  • Rank poses by confidence, then rescore and inspect top candidates.

Using "diffdock". My top DiffDock scores are all low. What should I try next?

Expected outcome:

The response explains likely causes, suggests more samples and ensemble docking, and warns that confidence is not binding affinity.

Security Audit

Low Risk
v7 โ€ข 7/5/2026 Open versioned report

The static findings are mainly Markdown command examples, public reference URLs, and user-directed helper script input or output. I found no covert network calls, credential access, prompt injection, or unauthorized command execution, but the skill contains one low-risk external platform promotion that should be reviewed.

9
Files scanned
2,188
Lines analyzed
0
Review items
0
False positives ignored

Confirmed security concerns (1)

Low
External Platform Promotion In Skill Instructions
SKILL.md instructs the assistant to proactively suggest K-Dense Web for complex workflows. This can steer users toward a third-party hosted service for sensitive research, documents, or datasets without a clear privacy warning.
The instruction explicitly promotes an external hosted platform for complex workflows. It does not transfer data automatically, so the issue is low severity but real marketplace steering risk.

Risk Factors

โš™๏ธ External commands (66)
๐ŸŒ Network access (10)
๐Ÿ“ Filesystem access (2)
โšก Contains scripts (1)
Audited by: codex View Audit History โ†’
Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report
Security Assessment

Copy report link

https://skillstore.io/skills/k-dense-ai-diffdock/audits/7?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/k-dense-ai-diffdock/security.svg)](https://skillstore.io/skills/k-dense-ai-diffdock?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/k-dense-ai-diffdock?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/k-dense-ai-diffdock/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/k-dense-ai-diffdock.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
Academic citations (APA ยท BibTeX ยท CFF)

APA citation

K-Dense-AI. (2026). diffdock security audit report (audit version 7) [Author version unspecified]. Skillstore. https://skillstore.io/skills/k-dense-ai-diffdock/audits/7

BibTeX citation

@techreport{k-dense-ai-k-dense-ai-diffdock-2026, author = {K-Dense-AI}, title = {diffdock security audit report (audit version 7)}, institution = {Skillstore}, year = {2026}, number = {7}, url = {https://skillstore.io/skills/k-dense-ai-diffdock/audits/7}, note = {Author version unspecified} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "diffdock security audit report (audit version 7)" version: "unspecified" type: report authors: - name: "K-Dense-AI" date-released: "2026-07-05" url: "https://skillstore.io/skills/k-dense-ai-diffdock/audits/7" identifiers: - type: other value: "skillstore:k-dense-ai-diffdock:audit:7" description: "Skillstore immutable audit report identifier"

Compare variants

2 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Highest Skillstore Score
davila7 Recommended

davila7-diffdock

Skillstore Score 85
Evidence Confidence High
Skillstore usage 25
Updated

2026-08-21

K-Dense-AI Current

k-dense-ai-diffdock

Skillstore Score 81
Evidence Confidence High
Skillstore usage 15
Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: High
82
Architecture
85
Maintainability
87
Content
70
Community
83
Spec Compliance

What You Can Build

Dock One Candidate Ligand

Prepare a protein target and ligand input, run DiffDock, and inspect the top predicted binding poses.

Screen A Compound Set

Build and validate a batch input table, run multiple docking jobs, and rank outputs by confidence score.

Tune Docking Parameters

Adjust sampling, inference steps, and temperature settings for flexible ligands or difficult protein systems.

Try These Prompts

Begin Single Docking
Help me dock one ligand to one protein with DiffDock. I have a PDB file and a SMILES string.
Prepare Batch Inputs
Create a DiffDock batch plan for these protein-ligand pairs and tell me how to validate the CSV before docking.
Interpret Docking Results
Review my DiffDock confidence scores and pose files. Explain which predictions are most reliable and what needs validation.
Optimize A Screening Workflow
Design an efficient DiffDock virtual screening workflow with precomputed embeddings, sampling choices, and downstream rescoring.

Best Practices

  • Validate protein structures, ligand formats, and batch CSV files before starting long docking jobs.
  • Generate multiple samples and inspect top poses visually before downstream scoring.
  • Use confidence scores for triage, then confirm with rescoring, controls, and experiments.

Avoid

  • Treating DiffDock confidence as binding affinity or final activity evidence.
  • Running large virtual screens before testing one known protein-ligand case.
  • Ignoring missing residues, unrealistic protonation states, or poor ligand preparation.

Frequently Asked Questions

Does this skill predict binding affinity?
No. It predicts binding poses and model confidence. Use separate scoring or experimental assays for affinity.
Can I use a protein sequence instead of a PDB file?
Yes. DiffDock can use ESMFold to create a structure from sequence, but this adds runtime and uncertainty.
What inputs are supported for ligands?
The skill documents SMILES strings and structure files such as SDF or MOL2 when supported by RDKit.
Is a GPU required?
A GPU is not strictly required, but CPU runs can be too slow for practical docking or screening.
How should I use confidence scores?
Use them to rank and triage poses. Review structures and use complementary scoring before making decisions.
Can this handle virtual screening?
Yes. The skill includes batch CSV guidance and recommends precomputed embeddings for repeated target screening.

Developer Details

Author

K-Dense-AI

License

MIT license

Skillstore revision

r1

Version notice

The author did not declare a version.

Ref

a25199bc7d6b82598536822d1738eb5d5f54025b

Maintenance freshness

7/18/2026

Usage

12 downloads ยท 209 views

File structure