Claude Managed Agents: sandbox self-hosted e tunnel MCP riportano gli agenti AI dentro il perimetro di sicurezza enterprise

Claude Managed Agents' MCP tunnels and sandboxes move credential control to the network boundary — a production fix for enterprise AI agent security.

Anthropic has quietly shipped two new infrastructure features for Claude agents: MCP tunnels and...

Claude Managed Agents: sandbox self-hosted e tunnel MCP riportano gli agenti AI dentro il perimetro di sicurezza enterprise