mmm-health-check
星标9
分支1
更新时间2026年5月12日 23:44
Trust-grade a Meridian MMM before acting on its outputs — convergence, prior-dominance, fit, model settings, and known failure patterns. Use when the user mentions "is the model good", "can I trust this MMM", "model diagnostics", "model health", "model audit", "prior posterior", "convergence check", "MMM validation", or "should I trust these numbers".
安装
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
文件资源管理器
2 个文件SKILL.md
readonly