How does the regression test verification work?

It follows a strict Red-Green cycle: proving the test fails with the bug, passes with the fix, fails again when the fix is temporarily reverted, and passes finally once restored.

Why does this skill forbid mocks in Tier 3?

To guarantee high-fidelity results, Tier 3 requires evidence from real browsers (like Puppeteer), real databases, and actual HTTP calls rather than simulated responses.

What are the three tiers of validation?

Tier 1 (Flow) checks for syntax and linting errors; Tier 2 (Artifact) ensures unit tests pass and builds succeed; Tier 3 (Functional) verifies real-world behavior using live systems without mocks.

Can I rely on previous test runs for verification?

No. The skill's 'Iron Law' mandates that all verification commands must be executed freshly within the current context to provide valid evidence for a completion claim.

Strict Verification Guard

Name: Strict Verification Guard
Author: krzemienski

bykrzemienski

•

Security & Testing

Enforces rigorous, evidence-based validation gates and fresh test execution before any coding task is marked as complete.

The Strict Verification Guard skill implements the 'Iron Law' of software development: no completion claims without fresh verification evidence. It moves beyond simple success messages by requiring Claude to execute and report on a three-tier validation framework covering Flow (syntax/linting), Artifacts (unit tests/builds), and Functional (real-world E2E) gates. By explicitly forbidding mocks in functional testing and requiring red-green regression cycles for bug fixes, this skill ensures that development output is actually functional, stable, and ready for production deployment.

Key Features

011 GitHub stars

02Three-tier validation gates (Flow, Artifact, and Functional)

03No-mocks enforcement for high-fidelity functional verification

04Standardized Red-Green-Red-Green cycle for regression testing

05Mandatory fresh command execution with quantitative reporting

06Automated detection of project-specific validation commands

Use Cases

01Validating complex refactors across compilation, unit, and E2E layers

02Preventing premature PR creation by requiring full test suite evidence

03Ensuring bug fixes are truly resolved through rigorous regression cycles

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add krzemienski/shannon-framework verification-before-completion

For use in Claude.ai and ChatGPT

Download Skill