ai-agent-guardrails

@goldenwing-360 · 收录于 1 周前

Apply safety controls when an LLM agent has authority to act on real systems. Covers blast-radius classification, dry-run-first patterns, out-of-band approval gates, scope locking, idempotency, kill switches, and rollback strategies. Invoke when designing an autonomous agent, when granting an LLM write access to production, or after an agent makes an unexpected change.

场景智能体上线前生产环境授权

适合你，如果你正在构建有真实系统操作权限的自主智能体

/ 下载安装

https://github.com/GoldenWing-360/claude-security-skills/tree/main/ai-agent-guardrails ↗

/ 通过 npx 安装 校验哈希

npx oh-my-skill add goldenwing-360/claude-security-skills/ai-agent-guardrails

/ 通过 bash 安装

curl -fsSL https://oh-my-skill.com/install.sh | bash -s -- goldenwing-360/claude-security-skills/ai-agent-guardrails

/ 已经装过?验证本机副本,不用重装

npx oh-my-skill verify goldenwing-360/claude-security-skills/ai-agent-guardrails

安装目标可用 --agent / --scope 或 --to 明确指定;省略时只会在唯一已存在的 agent 目录上自动选择,零命中或多命中会停止并提示。content_hash 缺失或不一致均拒装。

13GitHub stars

~1.9K上下文体积 · 单文件

索引托管

登录即可评论;带「已验证安装」的,是发布者名下有本店的安装或持有记录。

…