Agent Constitution Guard
Constitutional guardrails for AI agents — define immutable behavioral rules, permission boundaries, escalation policies, and safety interlocks.
0使用8点赞10浏览
这个技能会做什么
Constitutional guardrails for AI agents — define immutable behavioral rules, permission boundaries, escalation policies, and safety interlocks.
- 先确认目标、输入和成功标准
- 保留证据与无法确认的边界
- 输出可继续编辑的结构化结果