What happens if a claim is made without verification?

The skill triggers a violation recovery process, stopping the workflow and requiring the agent to run the missing commands and provide output before proceeding.

When does this skill automatically trigger?

It triggers whenever Claude attempts to claim work is complete, tests have passed, a bug is fixed, or when preparing to commit code.

Why is evidence-first verification important for AI agents?

It prevents 'silent failures' where an agent assumes code works without actually running tests, ensuring project integrity and human-level trust.

Does this skill prevent me from saying 'it should work now'?

Yes, 'should', 'appears to', and 'seems' are classified as Banned Phrases. The skill mandates that you provide evidence (output) rather than assumptions.

Evidence-First Verification Gate

Name: Evidence-First Verification Gate
Author: LerianStudio

byLerianStudio

•

Security & Testing

Enforces a mandatory command-first workflow that requires fresh verification evidence before any success claims or code commits.

This skill implements a strict quality gate designed to eliminate 'hallucinated' success in AI-driven development. By mandating a sequence of identifying verification commands, executing them, and documenting full output before claiming completion, it ensures every bug fix, test pass, or feature implementation is backed by concrete technical evidence. This skill is essential for maintaining high-trust workflows, preventing silent failures from reaching production, and enforcing professional software engineering discipline within Claude's environment.

Key Features

01Automated Red-Green regression testing validation

02Mandatory command-first verification protocol

0317 GitHub stars

04Violation recovery workflows for premature completion claims

05Detection and prevention of banned assumption-based phrases

06Fresh evidence enforcement for tests, builds, and linters

Use Cases

01Preventing false success reports during automated bug fixing

02Validating build integrity and linting after complex refactorings

03Ensuring 100% test pass rates before creating Pull Requests

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lerianstudio/ring verification-before-completion

For use in Claude.ai and ChatGPT

Download Skill