Tool Calling in Spring AI 2.0: A Composable, Agentic Architecture

A guide to Spring AI 2.0’s redesigned tool-calling system, which moves from model-specific implementations to a composable advisor-based architecture. It covers tool definition via annotations, the recursive ToolCallingAdvisor pattern, memory integration, scaling with ToolSearchToolCallingAdvisor for hundreds of tools, tool argument augmentation, MCP tool integration, custom advisors, user-controlled execution, and migration guidance from Spring AI 1.x.