References: The tool-use design pattern in depth

Source material

Source curriculum (structural mirror, cited as further study):
• Microsoft, "Tool Use Design Pattern" (AI Agents for Beginners, Lesson 04)
  Author: Microsoft Cloud Advocates
  Lesson page: https://github.com/microsoft/ai-agents-for-beginners/tree/main/04-tool-use
  License: MIT
  (This lesson's tool-definition content follows MS Lesson 04's building-blocks
   and function-calling sections.)
• Microsoft, "AI Agentic Design Principles" (AI Agents for Beginners, Lesson 03)
  Lesson page: https://github.com/microsoft/ai-agents-for-beginners/tree/main/03-agentic-design-patterns
  License: MIT
  (Broader agent design principles, transparency/control/consistency; cited as
   companion context, not as this lesson's tool-definition source.)
Clawdemy's lessons are original prose that follows the pedagogical arc of these
sources. We do not reproduce or transcribe them; we cite them as recommended
companions. All rights to the original materials remain with their authors.

Read this next

• Tool Use Design Pattern (Microsoft) by Microsoft Cloud Advocates. The practitioner version, with the building blocks of a tool definition and runnable function-calling samples. MIT-licensed. The closest source to this lesson’s content.
• AI Agentic Design Principles (Microsoft) by Microsoft Cloud Advocates. Higher-level design principles for agents (transparency, control, consistency in how an agent works with a user). Useful once you are designing the whole agent’s behavior, not just one tool.

Going deeper on defining tools

A short, durable list. Each is a primary source for how a real provider asks you to define tools, the practical form of everything in this lesson.

• Anthropic: Tool use with Claude. The provider docs for defining tools as schemas with names, descriptions, and typed parameters. Good descriptions and parameter docs are exactly what this page asks you to write.
• Model Context Protocol. An open standard for describing tools so they can be reused across models and applications. The same definition discipline, standardized.

Adjacent topics

Where this leads inside this track.

• Giving agents memory. The next lesson. So far every run started fresh; memory is how an agent holds on to information within a run and across runs, and how to decide what it should remember.
• How tool use turns a model into an agent. The previous lesson. The mechanism this lesson builds on: if “the model emits a tool call” is unclear, re-read it before going deeper on tool definitions.
• Building trustworthy agents. Later in the track. Tool definitions are also where reliability and safety begin; a tool defined with clear boundaries is one the model is less likely to misuse.