scikit-survival
Build survival models without leaking test data.
What it does
Guides the development and evaluation of survival-analysis pipelines with leakage-safe preprocessing and censoring-aware metrics.
When to use it
Use for right-censored survival modeling, model selection, time-dependent evaluation, and nonparametric competing-risk cumulative incidence with scikit-survival.
How to use it
It changes how the agent builds and evaluates survival-analysis code, requiring fold-contained preprocessing, independent evaluation, and training-only censoring estimates.
What you provide
- An existing project
Uses
Good to know
- advertises something the task does not need
What you need · 3
Executable examples require Python 3.11 or later.
Requires uv; the setup instructions use it to create an isolated environment and install the pinned scientific stack.
Access · 2
This skill
https://arxiv.org/abs/2609.00065
Fetches the publisher's paper metadata
scikit-survival
Installs the pinned survival-analysis stack