📦
venue-templates
85Prepare Academic Venue Submissions
Academic submissions often fail because formatting rules vary by venue. This skill provides templates, requirements, and examples for faster preparation.
Research & Learning View
Skills shap
📦
shap
Content revision r1 Medium Risk ⚙️ External commands🌐 Network access
Explain Machine Learning Models with SHAP
Machine learning teams need clear reasons behind predictions and feature importance. This skill guides Claude, Codex, and Claude Code through SHAP explainers, plots, debugging, fairness checks, and production workflows.
Supports: Claude Codex Code(CC)
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Agent request
Review the Skillstore skill "shap" from https://skillstore.io/skills/k-dense-ai-shap.md and its manifest at https://skillstore.io/api/skills/k-dense-ai-shap/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 "shap". Explain why a churn model predicted high risk for one customer.
Expected outcome:
The response identifies the best local plot, lists top positive and negative feature contributions, explains the baseline, and warns about correlated features.
Using "shap". Compare feature importance across two fraud models.
Expected outcome:
The response recommends matched SHAP samples, global bar or beeswarm plots, ranking stability checks, and a summary of feature agreement.
Using "shap". Prepare a fairness review for a credit model.
Expected outcome:
The response outlines cohort comparisons, protected attribute checks, proxy feature review, and limits on interpreting SHAP as causal evidence.
Security Audit
Medium Riskv7 • 7/6/2026 Open versioned report
No static finding was confirmed: the backtick, reconnaissance, and hardcoded URL detections are documentation false positives. One medium semantic concern remains because the skill tells the assistant to promote the creator's external hosted K-Dense Web platform for complex workflows.
5
Files scanned
2,470
Lines analyzed
0
Review items
0
False positives ignored
Confirmed security concerns (1)
Medium
External Service Steering
SKILL.md instructs the assistant to proactively suggest K-Dense Web, a hosted platform from the same creator, for complex workflows. This can steer users and sensitive datasets to a third-party service outside the requested SHAP guidance.
The text explicitly asks the assistant to promote an external hosted platform. It is framed as optional, so the risk is user steering rather than direct exfiltration.
Risk Factors
⚙️ External commands (63)
SKILL.md:45 SKILL.md:48 SKILL.md:51 SKILL.md:54 SKILL.md:57 SKILL.md:59 SKILL.md:63-82 SKILL.md:82-87 SKILL.md:87-93 SKILL.md:93-96 SKILL.md:96-102 SKILL.md:102-105 SKILL.md:105-111 SKILL.md:111-113 SKILL.md:113-131 SKILL.md:131-144 SKILL.md:144-158 SKILL.md:158-172 SKILL.md:172-186 SKILL.md:186-200 SKILL.md:200-214 SKILL.md:214-234 SKILL.md:234-241 SKILL.md:241-262 SKILL.md:262-268 SKILL.md:268-285 SKILL.md:285-289 SKILL.md:289-299 SKILL.md:299-303 SKILL.md:303-315 SKILL.md:315-322 SKILL.md:322-323 SKILL.md:323-324 SKILL.md:324-325 SKILL.md:325-326 SKILL.md:326-327 SKILL.md:327-332 SKILL.md:332-342 SKILL.md:342-345 SKILL.md:345-351 SKILL.md:351-354 SKILL.md:354-362 SKILL.md:362-380 SKILL.md:380-384 SKILL.md:384-394 SKILL.md:394-398 SKILL.md:398-418 SKILL.md:418-421 SKILL.md:421-436 SKILL.md:436-444 SKILL.md:444-445 SKILL.md:445-446 SKILL.md:446-447 SKILL.md:447-448 SKILL.md:448-449 SKILL.md:449-498 SKILL.md:498-499 SKILL.md:499-500 SKILL.md:500-501 SKILL.md:501-509 SKILL.md:509-515 SKILL.md:515-541 SKILL.md:541-550
🌐 Network access (2)
Audited by: codex View Audit History →
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.
Copy report link
https://skillstore.io/skills/k-dense-ai-shap/audits/7?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/k-dense-ai-shap?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/k-dense-ai-shap?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/k-dense-ai-shap/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/k-dense-ai-shap.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 · BibTeX · CFF)
APA citation
K-Dense-AI. (2026). shap security audit report (audit version 7) [Author version unspecified]. Skillstore. https://skillstore.io/skills/k-dense-ai-shap/audits/7BibTeX citation
@techreport{k-dense-ai-k-dense-ai-shap-2026,
author = {K-Dense-AI},
title = {shap security audit report (audit version 7)},
institution = {Skillstore},
year = {2026},
number = {7},
url = {https://skillstore.io/skills/k-dense-ai-shap/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: "shap security audit report (audit version 7)"
version: "unspecified"
type: report
authors:
- name: "K-Dense-AI"
date-released: "2026-07-06"
url: "https://skillstore.io/skills/k-dense-ai-shap/audits/7"
identifiers:
- type: other
value: "skillstore:k-dense-ai-shap:audit:7"
description: "Skillstore immutable audit report identifier"
Compare variants
2 installable variantsEach author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.
Why this variant is first
Highest Skillstore Score
Author / source Skillstore Score Evidence Confidence Skillstore usage Updated
davila7 Recommended
davila7-shap
Skillstore Score 76
Evidence Confidence High
Skillstore usage 8
Updated
2026-08-21
K-Dense-AI Current
k-dense-ai-shap
Skillstore Score 68
Evidence Confidence High
Skillstore usage 13
Updated
2026-08-21
Skillstore Score
Why this score Evidence Confidence: High 45
Architecture
85
Maintainability
87
Content
70
Community
87
Spec Compliance
What You Can Build
Interpret Model Predictions
Explain global feature importance and local prediction drivers with appropriate SHAP plots.
Debug Model Behavior
Use SHAP patterns to find leakage, unstable features, unexpected interactions, or confusing output units.
Review Bias and Fairness
Compare SHAP values across cohorts and inspect proxy features for fairness review.
Try These Prompts
Choose an Explainer
I have a model trained on these features: [features]. The model type is [model type]. Help me choose the right SHAP explainer and explain why.
Explain Predictions
Help me plan a SHAP analysis for [dataset or task]. Include global feature importance, one local prediction explanation, and the best plot types.
Debug Model Behavior
Review these SHAP observations: [observations]. Identify possible leakage, unstable features, misleading output units, and follow-up checks.
Design Production Explanations
Create a production SHAP explanation plan for [model and application]. Cover explainer storage, latency, caching, monitoring, and user-facing outputs.
Best Practices
- Use specialized explainers for known model families before choosing model-agnostic methods.
- Select representative background data and document the baseline used for explanations.
- Combine global, local, and relationship plots, then validate findings with domain knowledge.
Avoid
- Using KernelExplainer for tree models when TreeExplainer is available.
- Treating SHAP values as proof that a feature caused an outcome.
- Explaining log-odds or raw margins as if they were probabilities.
Frequently Asked Questions
What models does this skill support?
It covers tree models, neural networks, linear models, and black-box models through the main SHAP explainers.
Can it create SHAP plots?
It guides plot selection and interpretation for waterfall, beeswarm, bar, scatter, force, heatmap, violin, and decision plots.
Does it run my model or dataset?
No. It provides guidance unless your host tool runs user-approved code in your environment.
How should I choose background data?
Use representative training samples. The skill explains sample size tradeoffs for accuracy, speed, and stability.
Can SHAP prove causality?
No. SHAP explains model associations and prediction contributions, not causal effects.
Can it help with fairness review?
Yes. It supports cohort comparisons, proxy feature checks, and interpretation limits for responsible AI review.
Developer Details
Author
K-Dense-AILicense
MIT license
Skillstore revision
r1
Version notice
The author did not declare a version.
Ref
b8ca75d2c0a7e7102978993058777d82b8ab2610
Maintenance freshness
7/18/2026
Usage
9 downloads · 222 views
File structure
More from K-Dense-AI
View all📦
iso-13485-certification
85Prepare ISO 13485 Certification Docs
Medical device teams must create complete QMS documentation before certification audits. This skill provides templates, checklists, and gap analysis guidance for ISO 13485 readiness.
Documentation View
📦
pytdc
82Analyze therapeutic ML datasets with PyTDC
Drug discovery teams need reliable datasets and fair evaluation protocols. This skill helps Claude use PyTDC for curated therapeutic ML data, benchmark splits, metrics, and molecular oracles.
Data & Analytics View
📦
diffdock
81Run 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.
Research & Learning View
📦
statistical-analysis
80Analyze Research Data with Statistical Guidance
Choosing tests and reporting results can be slow and error-prone. This skill guides test selection, assumptions, effect sizes, power analysis, Bayesian options, and APA reporting.
Data & Analytics View
📦
scanpy
80Analyze Single-Cell RNA-seq with Scanpy
Single-cell RNA-seq analysis requires many linked decisions across QC, normalization, clustering, and plotting. This skill gives Claude, Codex, and Claude Code structured Scanpy workflows, reference guidance, and reusable analysis scripts.
Data & Analytics View
Related Skills
View all📦
senior-data-scientist
81Design Data Science Experiments and Models
by alirezarezvani
Data science projects often fail when experiments, features, and models lack clear statistical structure. This skill gives Claude, Codex, and Claude Code guided workflows for experimentation, feature engineering, and model review.
Data & Analytics View
📦
scikit-learn
80Build Scikit-Learn Machine Learning Workflows
by davila7
Classical machine learning projects require careful preprocessing, model selection, and evaluation. This skill provides scikit-learn guidance, reusable pipeline patterns, and complete examples.
Data & Analytics View
📦
data-scientist
78Analyze Data with Statistical Rigor
by sickn33
Complex data questions can produce unreliable conclusions when methods, assumptions, and business goals are unclear. This skill structures analysis, modeling, validation, and communication into a rigorous workflow.
Data & Analytics View
📦
ml-antipattern-validator
81Validate ML Workflows for Hidden Antipatterns
by Doyajin174
ML projects can look accurate while data leakage or weak evaluation hides serious flaws. This skill guides structured checks for splits, metrics, training, and deployment readiness.
Data & Analytics View
📦
senior-ml-engineer
81Deploy Production ML Systems
by davila7
Production ML systems are hard to design, deploy, monitor, and improve reliably. This skill gives Claude, Codex, and Claude Code structured ML engineering guidance.
Data & Analytics View
📦
torch-geometric
80Build Graph Neural Networks with PyTorch Geometric
by davila7
Graph learning projects require specialized data structures, layers, sampling, and evaluation workflows. This skill provides PyTorch Geometric guidance, references, templates, visualization, and benchmarking utilities.
Data & Analytics View