Google Deepmind adds background execution and MCP support to Gemini API managed agents

Google DeepMind has updated its Gemini API managed agents to support asynchronous background execution and direct connections to Model Context Protocol servers. These changes allow AI agents to handle longer tasks and retrieve external data more efficiently without requiring constant user input. By enabling persistent state during credential refreshes and the integration of custom functions with existing tools, the update improves the reliability of automated workflows. These technical improvements make it easier for developers to build agents that perform complex, multi-step actions across various software environments.