How does the reliability tracking work?

The skill tracks pass/fail history in a local YAML file, calculating a reliability score (0.0 to 1.0) to ensure flaky tests don't slow down your Tier 1 'Fast' runs.

Can I use this for non-Python projects?

Currently, this skill is specifically optimized for Python environments using Pytest and standard Python import structures.

When should I run the full test suite instead of a smart test?

You should run the full suite during major refactoring, database schema changes, global configuration updates, or as a final check before merging into the main branch.

What are the different testing tiers?

Tier 1 runs direct unit tests in under 1 minute; Tier 2 includes transitive dependencies and integration tests; Tier 3 runs the entire suite for final validation.

How does Smart Test Selection know which tests to run?

It performs an import dependency analysis on your Python files, building a map between source code and the tests that import or depend on those modules.

Smart Test Selection

Name: Smart Test Selection
Author: rysweet

byrysweet

•

Security & Testing

Optimizes development cycles by intelligently selecting and running only the tests affected by your recent code changes.

Smart Test Selection enhances the developer workflow by mapping source code changes to relevant tests through deep import dependency analysis. It implements a sophisticated three-tier testing strategy—Fast, Impacted, and Full Suite—to provide immediate feedback while tracking test reliability to automatically filter out flaky results. This skill is essential for developers working on large codebases who need to reduce CI/CD overhead and minimize local feedback loops during rapid iteration without compromising code quality.

Key Features

01Impacted test mapping via Python import dependency analysis

02Automated reliability tracking to identify and isolate flaky tests

0317 GitHub stars

04Seamless integration with Git diffs and Pytest markers

05Three-tier testing strategy (Fast < 1min, Impacted < 5min, Full)

06Dynamic generation of ready-to-run terminal commands

Use Cases

01Identifying and quarantining unstable tests that disrupt CI/CD pipelines

02Reducing pre-commit execution time by running only unit tests for modified files

03Validating complex refactors by identifying and running transitive dependency tests

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rysweet/amplihack smart-test

For use in Claude.ai and ChatGPT

Download Skill