Agentic Loops and Tool Use

aid: context-engineering:agent-loops
name: Agentic Loops and Tool Use
description: 'Agentic loops are iterative reasoning patterns in which an LLM plans, calls tools, observes
  results, and refines its plan. Tool design is a central context engineering concern: tools must be token-efficient,
  have minimal overlap, and include clear, motivating descriptions.'
humanURL: https://docs.anthropic.com/en/docs/build-with-claude/tool-use/overview
baseURL: https://docs.anthropic.com
tags:
- Agents
- Function Calling
- ReAct
- Tool Use
properties:
- type: Documentation
  url: https://docs.anthropic.com/en/docs/build-with-claude/tool-use/overview
- type: Reference
  url: https://platform.openai.com/docs/guides/function-calling
- type: Reference
  url: https://arxiv.org/abs/2210.03629
x-features:
- Tool definitions with JSON Schema arguments
- Iterative plan-act-observe loops
- Parallel tool invocation
- Server- and client-side tool execution
x-useCases:
- Building task-completing AI agents
- Wiring LLMs to internal APIs and databases
- Decomposing complex problems with sub-tools