yes-zh

What problem does it solve?

This Skill acts as an AI governance engine, ensuring AI assistants operate safely, evidence-based, and thoroughly by enforcing strict rules and preventing common AI bad habits like guessing, deflecting, and superficial fixes.

Core Features & Use Cases

• Enforces Evidence Rules: Prevents AI from making claims without data, requiring verification before conclusions.
• Implements Safety Gates: Mandates backups before modifications, impact analysis, and pre-deployment checks.
• Detects and Corrects Bad Habits: Catches and corrects seven common AI shortcuts, including guessing, deflecting, and superficial fixing.
• Structured Debugging: Guides AI through a systematic debugging process with increasing levels of self-inspection upon repeated failures.
• Use Case: When an AI coding assistant is asked to fix a bug, this Skill ensures it first backs up the relevant files, checks the blast radius of the change, verifies the fix with tests, and then checks for related issues before confirming completion.

Quick Start

Use the yes-zh skill to ensure the AI follows evidence-based rules and safety gates before making any code changes.