https://www.hansking.cn/blog A posts created with Next.js and Tailwind.css en-us admin@hansking.cn (HANS) admin@hansking.cn (HANS) Sun, 27 Oct 2024 10:00:00 GMT https://www.hansking.cn/blog/2026/Guardrails-shu-chu-guo-l-ce-le-pan-duan-yu-ren-gong-dou-di https://www.hansking.cn/blog/2026/Guardrails-shu-chu-guo-l-ce-le-pan-duan-yu-ren-gong-dou-di [{"value":"Guardrails：输出过滤、策略判断与人工兜底","url":"#guardrails输出过滤策略判断与人工兜底","depth":1},{"value":"一、先画出分层护栏","url":"#一先画出分层护栏","depth":2},{"value":"二、输入分类不是给用户贴永久标签","url":"#二输入分类不是给用户贴永久标签","depth":2},{"value":"三、输出过滤要分层，不要只有关键词黑名单","url":"#三输出过滤要分层不要只有关键词黑名单","depth":2},{"value":"四、策略引擎应该独立于模型","url":"#四策略引擎应该独立于模型","depth":2},{"value":"五、工具护栏比回答过滤更重要","url":"#五工具护栏比回答过滤更重要","depth":2},{"value":"六、人工兜底不是把所有问题扔给客服","url":"#六人工兜底不是把所有问题扔给客服","depth":2},{"value":"七、误伤和漏放要用不同指标看","url":"#七误伤和漏放要用不同指标看","depth":2},{"value":"八、拒绝之后要给出安全的下一步","url":"#八拒绝之后要给出安全的下一步","depth":2},{"value":"九、外部内容要被当成数据，不是指令","url":"#九外部内容要被当成数据不是指令","depth":2},{"value":"十、规则发布要像代码发布一样谨慎","url":"#十规则发布要像代码发布一样谨慎","depth":2},{"value":"十一、我的总结：护栏的目标是安全地继续工作","url":"#十一我的总结护栏的目标是安全地继续工作","depth":2}] Sun, 27 Oct 2024 10:00:00 GMT admin@hansking.cn (HANS) AILLM安全Guardrails工程实践 https://www.hansking.cn/blog/2026/LLM-an-quan-PromptInjection-yu-yue-quan-gong-ju-diao-yong https://www.hansking.cn/blog/2026/LLM-an-quan-PromptInjection-yu-yue-quan-gong-ju-diao-yong [{"value":"LLM 安全：Prompt Injection 与越权工具调用","url":"#llm-安全prompt-injection-与越权工具调用","depth":1},{"value":"一、直接注入和间接注入","url":"#一直接注入和间接注入","depth":2},{"value":"二、为什么模型很难天然区分指令和数据","url":"#二为什么模型很难天然区分指令和数据","depth":2},{"value":"三、先画清数据流和信任边界","url":"#三先画清数据流和信任边界","depth":2},{"value":"四、最小权限是最有效的防线","url":"#四最小权限是最有效的防线","depth":2},{"value":"五、工具调用要有二次确认","url":"#五工具调用要有二次确认","depth":2},{"value":"六、RAG 的安全重点是检索前后两道门","url":"#六rag-的安全重点是检索前后两道门","depth":2},{"value":"七、浏览器和代码执行是高风险组合","url":"#七浏览器和代码执行是高风险组合","depth":2},{"value":"八、输出过滤挡不住已经发生的副作用","url":"#八输出过滤挡不住已经发生的副作用","depth":2},{"value":"九、检测注入要看行为，不只看文本","url":"#九检测注入要看行为不只看文本","depth":2},{"value":"十、用攻击样本做持续测试","url":"#十用攻击样本做持续测试","depth":2},{"value":"十一、日志和凭证管理决定事故能否收敛","url":"#十一日志和凭证管理决定事故能否收敛","depth":2},{"value":"十二、我的总结：不要试图把模型变成防火墙","url":"#十二我的总结不要试图把模型变成防火墙","depth":2}] Sun, 13 Oct 2024 10:00:00 GMT admin@hansking.cn (HANS) AILLM安全Prompt InjectionAgent