video1.0<iframe src="https://www.loom.com/embed/7cac8fbb58d340d49cf36fbcba50db02" frameborder="0" width="1280" height="960" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>9601280Loomhttps://www.loom.com9601280https://cdn.loom.com/sessions/thumbnails/7cac8fbb58d340d49cf36fbcba50db02-1322530bf7d898c9.gif298.484This Loom demonstrates a project that uses dependency-aware access control for SQL and context sharing across AI agents. It explains how adaptive SPY searches are blocked by table dependencies, so delete operations fail when related dependencies exist. The author shows building a context layer that evaluates relationships and retrieves data through scripted execution, then confirms the agent can access related tables such as stockholder and account views. Finally, it shows that other agents can be granted access through a LCP server while external agents do not get uncontrolled access to the context layer.