AI Agent Authorization: A Least-Privilege Operating Model

Описание и применение

<p><strong>Identity and authorization are separate checks.</strong> A signed workload identity can tell the resource server who is calling; it cannot by itself establish that a particular principal delegated a write on a particular table. A robust architecture represents both identities and the right exercised in each operation. This distinction becomes critical with ephemeral instances, multi-tenant tools and subagents.</p> <p><strong>Represent grants as action tuples.</strong> A practical tuple contains principal, authenticated agent, task, target tenant/object, operation, authorized tool, c

Задача

<p><strong>Identity and authorization are separate checks.</strong> A signed workload identity can tell the resource server who is calling; it cannot by itself establish that a particular principal delegated a write on a particular table. A robust architecture represents both identities and the right exercised in each operation. This distinction becomes critical with ephemeral instances, multi-tenant tools and subagents.</p> <p><strong>Represent grants as action tuples.</strong> A practical tuple contains principal, authenticated agent, task, target tenant/object, operation, authorized tool, c

Результат

<p><strong>Identity and authorization are separate checks.</strong> A signed workload identity can tell the resource server who is calling; it cannot by itself establish that a particular principal delegated a write on a particular table. A robust architecture represents both identities and the right exercised in each operation. This distinction becomes critical with ephemeral instances, multi-tenant tools and subagents.</p> <p><strong>Represent grants as action tuples.</strong> A practical tuple contains principal, authenticated agent, task, target tenant/object, operation, authorized tool, c

Доказательства

AI Agent Authorization: A Least-Privilege Operating Model

Источник