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

•

安全与测试

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

关于

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.

主要功能

Automated Red-Green regression testing validation
Mandatory command-first verification protocol
17 GitHub stars
Violation recovery workflows for premature completion claims
Detection and prevention of banned assumption-based phrases
Fresh evidence enforcement for tests, builds, and linters

使用场景

Preventing false success reports during automated bug fixing
Validating build integrity and linting after complex refactorings
Ensuring 100% test pass rates before creating Pull Requests

关于

主要功能

Automated Red-Green regression testing validation
Mandatory command-first verification protocol
17 GitHub stars
Violation recovery workflows for premature completion claims
Detection and prevention of banned assumption-based phrases
Fresh evidence enforcement for tests, builds, and linters

使用场景

Preventing false success reports during automated bug fixing
Validating build integrity and linting after complex refactorings
Ensuring 100% test pass rates before creating Pull Requests