skills.rest
kriscendobot
kriscendobotCOMMUNITY
Explore all skills by this author →
GitHub

regression-evidence

Validate new test cases by enforcing regression-evidence protocols during code review.

3|2|Updated May 12, 2026
Search Tags:#testing#quality assurance#code review#refactoring#verification#regression
One-click install
npx skills add https://github.com/kriscendobot/garden --skill regression-evidence
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: regression-evidence
Source: https://github.com/kriscendobot/garden/tree/main/skills/regression-evidence
Command: npx skills add https://github.com/kriscendobot/garden --skill regression-evidence

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill addresses the issue of unreliable test suites by ensuring that every new test is load-bearing and capable of catching actual regressions, preventing silent failures in your codebase.

Core Features & Use Cases

  • Regression Verification: Enforces a discipline of temporarily breaking code to confirm that tests fail as expected, ensuring they are not merely passing by coincidence.
  • Equivalence Testing: Provides a framework to verify that refactored code or dependency upgrades maintain identical behavior to previous versions.
  • Differential Probing: Uses local fakes to compare wire-level behavior across dependency versions, ensuring major upgrades do not introduce silent breaking changes.

Quick Start

Ask the garden to verify the regression evidence for the latest test by running the prover seat against the current diff.

Frequently Asked Questions about regression-evidence

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify that new test cases actually catch code regressions?▼

To verify regression evidence, you must temporarily break the code paths to confirm the tests fail as expected, ensuring they are load-bearing and not passing by coincidence. This failure-mode experiment proves test integrity.

What is regression evidence in software testing and when do I need it?▼

Regression evidence is the substantiation that a test catches actual regressions by demonstrating failure on broken code. You need it during code review to prevent silent failures and ensure every new test is load-bearing.

How do I test equivalence when refactoring code or upgrading dependencies?▼

You test equivalence during refactoring by using differential probing with local fakes to compare wire-level behavior across dependency versions, ensuring the refactored code maintains identical behavior to previous versions.

Can I use this regression verification protocol for major dependency upgrades?▼

Yes, the regression verification protocol supports major dependency upgrades by using differential probing to compare wire-level behavior across versions, ensuring upgrades do not introduce silent breaking changes.

Why does my test suite pass but still allow silent failures in the codebase?▼

Your test suite allows silent failures because tests may be passing by coincidence rather than validating broken code paths. Without enforcing failure-mode experiments, tests are not load-bearing and cannot catch actual regressions.

Related Skills

Explore More Software Engineering Skills→
forcedotcomforcedotcom
803

running-apex-tests

Execute Salesforce Apex unit tests with coverage analysis and failure diagnosis.

Official
Advanced
ibalasiteibalasite

investigate

Diagnose software bugs and runtime errors with root cause analysis and regression tests.

Community
Advanced
rortan134rortan134
26

find-bugs-diff

Identify bugs, security vulnerabilities, and code quality issues in source code changes.

Community
Advanced
adityaarsharmaadityaarsharma
2

orbit-qa-coverage

Analyze WordPress plugin code coverage during PHPUnit test execution.

Community
Advanced