Runtime authorization for AI agents. Before Cursor's agent reads, writes, deletes, exports or sends data, ask Vulnify and get ALLOW, REVIEW or BLOCK with a risk score and reasons. Sign in with Vulnify…