How does the skill know which tests to run?

The skill performs import dependency analysis to map source files to their corresponding test files, identifying direct matches as well as transitive dependencies.

When should I avoid using Smart Test Selection?

Avoid using it during major refactoring, database schema migrations, or as the final check before merging to main, where a full suite run is required.

How is test reliability tracked?

The skill maintains a local YAML database that records pass/fail rates, allowing it to exclude flaky tests from high-speed feedback loops.

Does it support languages other than Python?

The current implementation is optimized specifically for Python environments and integrates directly with the pytest framework.

What are the three testing tiers?

Tier 1 (Fast) targets unit tests under 1 minute; Tier 2 (Impacted) runs all affected tests including integrations under 5 minutes; Tier 3 is the full suite run.

Smart Test Selection

Name: Smart Test Selection
Author: rysweet

byrysweet

•

Security & Testing

Optimizes developer feedback loops by intelligently selecting and running only the tests affected by specific code changes.

Smart Test Selection is a production-ready utility for Claude Code that minimizes test execution time by analyzing Python import dependencies and git diffs to identify impacted code paths. It implements a sophisticated three-tier testing strategy—ranging from ultra-fast pre-commit checks to comprehensive full-suite runs—while tracking test reliability to automatically filter out flaky tests. By providing copy-paste ready pytest commands and time estimates, this skill helps developers maintain high velocity in large codebases without sacrificing testing rigor or waiting for unnecessary CI cycles.

Key Features

0117 GitHub stars

02Seamless integration with pytest, git workflows, and pre-commit hooks.

03Visual test analysis reports with estimated execution times for each tier.

04Automated reliability tracking and identification of flaky tests.

05Intelligent mapping of source code to tests via import dependency analysis.

06Three-tier testing strategy (Fast, Impacted, Full Suite) for optimized feedback.

Use Cases

01Identifying impacted integration tests for a specific Pull Request to save CI time.

02Detecting and quarantining flaky tests based on historical pass/fail data.

03Running ultra-fast, high-reliability unit tests locally before every code commit.

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