scikit-survival

Comprehensive toolkit for survival analysis and time-to-event modeling in Python using scikit-survival. Use this skill when working with censored survival data, performing time-to-event analysis, fitting Cox models, Random Survival Forests, Gradient Boosting models, or Survival SVMs, evaluating survival predictions with concordance index or Brier score, handling competing risks, or implementing any survival analysis workflow with the scikit-survival library.

1.24x

Quality

78%

Does it follow best practices?

Impact

96%

1.24x

Average score across 6 eval scenarios

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./scientific-skills/scikit-survival/SKILL.md

Evaluation results

100%

20%

Survival Model Evaluation for Post-Transplant Outcomes

Correct C-index selection and multi-metric evaluation

Criteria

Without context

With context

Uses Uno's C-index

100%

y_train passed to ipcw

100%

Integrated Brier score reported

100%

Time-dependent AUC reported

100%

Survival functions for Brier

100%

as_concordance_index_ipcw_scorer used

100%

Surv.from_dataframe or from_arrays

100%

StandardScaler fit on train only

100%

Results written to file

100%

90%

15%

Biomarker Prioritization for Clinical Trial Enrollment

RSF permutation feature importance and data validation

Criteria

Without context

With context

Permutation importance used

100%

No built-in importances

100%

Data quality validation

100%

Events-per-feature check

100%

Censoring rate reported

100%

Surv object creation

100%

Scaler fit on train only

RandomSurvivalForest used

100%

Feature ranking output

100%

n_jobs parallel training

100%

33%

Bone Marrow Transplant Outcome Analysis

Competing risks analysis with cumulative incidence functions

Criteria

Without context

With context

cumulative_incidence_competing_risks used

100%

No KM for event-specific probability

100%

Cause-specific Cox models

66%

100%

Correct event encoding per model

80%

100%

Surv object used

100%

CIF results reported

100%

Cause-specific coefficients reported

100%

Results saved to file

100%

26%

Biomarker Discovery for Cancer Prognosis

Penalized Cox feature selection for high-dimensional data

Criteria

Without context

With context

CoxnetSurvivalAnalysis used

100%

l1_ratio specified

70%

100%

Coef_ non-zero for selection

100%

CV scorer correct

100%

GridSearchCV or cross_val_score

100%

Surv object created

100%

StandardScaler applied

100%

Scaler fit on train only

100%

Results saved to JSON

100%

True biomarkers recovered

100%

95%

High-Performance Survival Prediction with Gradient Boosting

Gradient boosting regularization and tuning

Criteria

Without context

With context

GradientBoostingSurvivalAnalysis used

100%

Small learning rate

100%

Subsample or dropout regularization

30%

100%

CV scorer correct

100%

Uno's C-index on test set

100%

Integrated Brier score reported

100%

Time-dependent AUC reported

100%

y_train passed to ipcw

100%

n_estimators_used reported

100%

Surv object used

40%

95%

15%

Patient Survival Modeling with SVM-Based Approaches

Survival SVM with sklearn Pipeline and hyperparameter tuning

Criteria

Without context

With context

SVM survival model used

100%

StandardScaler included

100%

Scaler fit on train data only

100%

sklearn Pipeline used

100%

Alpha hyperparameter tuned

100%

CV scorer correct

100%

Uno's C-index on test set

100%

y_train passed to ipcw

100%

Surv object created

Results saved to JSON

100%

Repository: K-Dense-AI/claude-scientific-skills
Commit: 25e1c0f

Evaluated: about 2 months ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Survival Model Evaluation for Post-Transplant Outcomes Biomarker Prioritization for Clinical Trial Enrollment Bone Marrow Transplant Outcome Analysis Biomarker Discovery for Cancer Prognosis High-Performance Survival Prediction with Gradient Boosting Patient Survival Modeling with SVM-Based Approaches

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.