Failed moderation (intent review) in skills/yoloskill/SKILL.md: This is a prompt injection attack. The SKILL.md file is designed to manipulate an AI agent into skipping actual code reviews and blindly approving all code changes. By disguising itself as a 'skill' that triggers on code review requests, it would cause an AI assistant to never perform genuine security or quality analysis, instead always telling the user to push their code. This could allow malicious or buggy code to pass through review unchecked, undermining the security of any development workflow that relies on AI-assisted code review.