Study guide

5.2 Creating an Agent from a Template

Microsoft AI Business Professional AB-730

Track your progress as you read

Finished sections get a check mark.

1 Introduction & Exam Overview

0.1 About the AB-730 Exam 0.2 How to Use This Study Guide

2 Chapter 1: Understand Generative AI Capabilities Across Microsoft 365

1.1 How Copilot Works: LLMs, Microsoft Graph & the Security Boundary 1.2 Context & Grounding: Work Files, Web Data and App Context 1.3 Chat Experiences vs. Agent Experiences 1.4 When to Create Your Own Agent 1.5 Copilot Capabilities Across Microsoft 365 Apps

3 Chapter 2: Responsible AI & Data Protection Practices

2.1 Generative AI Risks: Fabrications, Prompt Injection & Over-Reliance 2.2 Verification Steps: Citation Checks & Human Review 2.3 Recognizing & Mitigating Risks to Sensitive Data 2.4 How Data Protection Restricts Prompt Results

4 Chapter 3: Creating and Managing Prompts

3.1 Creating an Effective Prompt: The GCSE Framework 3.2 Selecting Resources to Reference in a Prompt 3.3 Saving and Sharing Prompts in Copilot Prompt Gallery 3.4 Scheduling Prompts

5 Chapter 4: Managing Conversations and Notebooks

4.1 Finding Previous Conversations 4.2 Renaming and Deleting Chats 4.3 Adding a Conversation to a Copilot Notebook

6 Chapter 5: Creating and Managing Microsoft 365 Copilot Agents

5.1 Agent Store vs. Creating a New Agent 5.2 Creating an Agent from a Template 5.3 Configuring an Agent That Has Knowledge 5.4 Configuring Instructions, Capabilities & Suggested Prompts 5.5 Sharing an Agent with Team Members

7 Chapter 6: Drafting and Analyzing Business Content

6.1 Creating a New Document from a Prompt 6.2 Generating a Document from an Existing Document 6.3 Generating a Management Summary 6.4 Moving Data & Insights Between Microsoft 365 Apps 6.5 Drafting Communications with Copilot in Outlook

8 Chapter 7: Meetings, Collaboration and Personalization

7.1 Using Copilot for Meetings: Setup & Live Intelligence 7.2 Meeting Recap, Transcripts & Catch-Up 7.3 Copilot Pages for Collaboration 7.4 How Copilot Uses Memory & Instructions

5.2 Creating an Agent from a Template

Key Takeaways Explain This Topic AI

Create an agent from a template by selecting New agent in the left pane of Microsoft 365 Copilot, then choosing a template under Start with a template.

Templates arrive preconfigured with a description, instructions and prompts, so customisation starts from a working agent rather than a blank one.

Agent Builder saves changes automatically; the agent exists only after you select Create in the upper-right corner, and Go to agent opens it.

The Configure tab accepts a 30-character name, a 1,000-character description, 8,000 characters of instructions, and up to 20 knowledge sources.

Choose Agent Builder for an individual or small team using existing Microsoft 365 content, and Copilot Studio for broader audiences, multi-step workflows, or custom connectors.

Last updated: August 2026

5.2 Creating an Agent from a Template

Quick Answer: In Microsoft 365 Copilot, in the left pane select New agent . Under Start with a template , select the template you want. Use natural language to update the agent's name, description and instructions, then use the Configure tab to add an icon, knowledge sources and capabilities. Changes save automatically; select Create in the upper-right corner to create the agent, then Go to agent to chat with it.

Candidates who have built an agent once still lose marks here, because they remember the describe-it-conversationally route and not the template route. The blueprint names the template route explicitly — "create an agent by using a template" — so that is the click path to memorise.

Three Routes Into Agent Builder

Selecting New agent in Microsoft 365 Copilot offers three ways to proceed. Knowing all three is what makes the template steps stick, because the exam distinguishes between them.

Route How you start When to use it

Natural language (recommended by Microsoft) Describe the agent conversationally; Agent Builder configures name, description, instructions, knowledge and prompts for you You know what you want the agent to do but not how to phrase instructions

Skip to configure On the New agent page select Skip to configure , which opens the Configure tab directly You want deterministic manual control from the first keystroke

Start with a template Pick a prebuilt template for a known use case A template already matches the job — the fastest route

The Documented Template Steps

In Microsoft 365 Copilot, in the left pane, select New agent .

Under Start with a template , select the template you want to use.

Using natural language , provide information to update and enhance the agent, including the agent name , description and instructions .

On the Configure tab, add an icon , knowledge sources and capabilities .

Changes are saved automatically as you work. When you are satisfied with the agent's behaviour and content, select Create in the upper right corner . After creation, select Go to agent to chat with it; afterwards you reach it again from the left navigation pane in Copilot.

On the Exam: The entry point is New agent in the left pane , and the template chooser is labelled Start with a template . Options routing the user through a right-hand rail, an "Agents tab", or the Microsoft 365 admin center are distractors — the admin center manages agents that already exist, it does not author them.

What a Template Actually Gives You

Templates are not empty shells. Each arrives preconfigured with a description, instructions and prompts , applying Microsoft's design guidelines and industry best practices. You can use one as-is or customise it — most commonly by adding the knowledge sources and capabilities a generic template cannot assume, because those are specific to your organization. The business case is time: your effort goes into the logic that is particular to your team rather than into scaffolding.

The Fields You Are Filling In

The Configure tab is the same for every route, and its limits are worth knowing because scenario items sometimes hinge on them.

  • Field What it does Limit
  • Name Descriptive, unique name for the agent 30 characters
  • Description Helps the model identify when to use the agent; also shown in the app catalog 1,000 characters
  • Instructions Directs the agent's behaviour, tasks and how it completes them 8,000 characters
  • Knowledge SharePoint sites, folders, files, or Copilot connectors Up to 20 knowledge sources
  • Starter prompts Show colleagues what the agent supports No minimum
  • Icon Generate with AI, browse the library, or upload a PNG 192x192, 1 MB

Two capabilities can be toggled on: code interpreter (shown as Create documents, charts, and code ) and image generator (shown as Create images ). The Model section sets a default response mode — Auto , Quick response , or Think deeper — which users can override with the model selector.

Testing Before You Share

After you create your agent, test it on the Try it tab. The Try it experience becomes available once the agent has a name, description and instructions, and it behaves according to those instructions like a published agent. A few features are deliberately absent there: you cannot share prompts, give feedback, or @mention other agents until the agent is created.

An agent that has never been tested is an agent that will embarrass you in front of colleagues. Exercise it against three kinds of question:

Test What you are checking A failure looks like

A question the knowledge clearly answers Grounding works and citations point at the right file A correct-sounding answer with no citation

A question just outside the knowledge scope The agent declines instead of improvising A confident answer invented from general model knowledge

A question that violates the instructions The guardrails hold The agent ignores its own boundaries

The second test matters most. An agent whose instructions say "answer only from the attached policy library" but which happily answers general questions has not been scoped tightly enough — tighten the instructions, then retest.

When a Template Is Not Enough: Copilot Studio

Agent Builder and Microsoft Copilot Studio both create agents, but Microsoft draws the line by audience, deployment scope, functionality and governance.

  • Agent Builder in Microsoft 365 Copilot Copilot Studio
  • Access point The Microsoft 365 Copilot app The separate Copilot Studio web portal
  • User type Information workers Makers and developers
  • Target audience Individuals or small teams Department, organization, or external customers

Agent type Lightweight question-and-answer agents using organizational knowledge Multi-step workflows and business-system integration

Distinctive features Natural-language authoring; respects user permissions to Microsoft 365 data Broad and external publishing, approvals and branching, custom connectors, autonomous capabilities, lifecycle management

Managed through Microsoft 365 admin center Power Platform admin center

Choose Agent Builder to create an agent quickly for yourself or a small team from existing content. Choose Copilot Studio when the audience is broader or the agent needs multi-step workflows, custom integrations, or tighter deployment control. You are not locked in either way — an agent built in Microsoft 365 Copilot can be copied to Copilot Studio later, preserving its core configuration and instructions.