Coinworld News reported that in response to key security issues, Tencent Cloud AI Agent Security Center launched "Credential Sandbox," with the principle of "Agent Never Holds Keys." According to the introduction, through the Credential Sandbox, Agent can complete all cloud API calls without holding any keys. Each operation is fully auditable throughout the process, and administrators can adjust the ability boundaries of Agent at any time——granting permissions without providing keys.
查看原文