Once individual tools are developed, the challenge becomes enabling an LLM agent to effectively choose the right tool for a given situation and to coordinate multiple tools for more complex operations. This chapter focuses on exactly that: how LLM agents select and sequence tools. We will examine strategies for agent-driven tool selection, the design of multi-step tool execution flows, and techniques for managing dependencies between tool calls. You will also learn to implement conditional logic for tool usage and address the coordination of both sequential and parallel tool execution.