新发现
prismorsec/prismor v1.68.3 — 智能体工具调用管控层
prismorsec/prismor 是一个自托管的运行时安全工具,用于观察 AI 智能体的工具调用,并对其批准或阻止。
prismorsec/prismor 是一个自托管的运行时安全工具,用于观察 AI 智能体的工具调用,并对其批准或阻止。
GitHub 项目说明: Self-hosted runtime control plane for AI agents. Observe or HITL approve or Block rogue tool calls before it executes: secret leaks, prompt injection, supply chain etc in a local dashboard. Agent agnostic (Claude, codex, langchain etc.)
在 GitHub 查看 ↗prismorsec/prismor 是一个自托管的运行时安全工具,用于观察 AI 智能体的工具调用,并对其批准或阻止。
没有符合条件的动态。