Study guide
13.1 Building Prompts from Templates or from Blank
Microsoft Certified: Intelligent Applications Builder Associate AB-410
Track your progress as you read
Finished sections get a check mark.
1 Introduction
1.1 About the AB-410 Exam 1.2 How to Use This Study Guide
2 Chapter 2: Designing Power Platform Solutions with AI-Enabled Tools
2.1 Analyzing Requirements and Identifying Components 2.2 Evaluating Built-in Agents for Business Solutions 2.3 Recommending Extensibility Options
3 Chapter 3: Environments and ALM Strategy
3.1 Recommending Environment Types 3.2 Solution and Application Lifecycle Management Strategy
4 Chapter 4: Dataverse Tables and Columns
4.1 Creating and Editing Tables in the Data Workspace 4.2 Configuring Table Properties 4.3 Creating and Modifying Columns
5 Chapter 5: Relationships, AI Data Features, and Data Security
5.1 Configuring Table Relationships 5.2 Configuring Prompt Columns and Row Summaries 5.3 Configuring Public Views and Main Forms 5.4 Configuring Dataverse Security
6 Chapter 6: Model-Driven App Forms and Views
6.1 Creating and Configuring Forms for Model-Driven Apps 6.2 Creating and Configuring Views for Model-Driven Apps 6.3 Creating Generative Pages with Natural Language
7 Chapter 7: Composing, Securing, and Visualizing Model-Driven Apps
7.1 Composing a Model-Driven App 7.2 Configuring Access to Model-Driven Apps, Forms, and Views 7.3 Model-Driven App Charts and Dashboards
8 Chapter 8: Creating and Designing Canvas Apps
8.1 Creating a Canvas App Using Data 8.2 Designing for Accessibility, Performance, Responsiveness, and Usability 8.3 Automating Business Processes from Canvas Apps
9 Chapter 9: Canvas App Components, Variables, and Error Handling
9.1 Reusable Components, Named Formulas, and User-Defined Functions 9.2 Managing Variables and Collections 9.3 Implementing Error Handling in Canvas Apps
10 Chapter 10: Testing Canvas Apps and Creating Agents from Apps
10.1 Testing Canvas Apps with Monitor 10.2 Creating a Copilot Studio Agent from a Canvas App
11 Chapter 11: Cloud Flow Triggers, Connectors, and Approvals
11.1 Recommending Cloud Flow Triggers 11.2 Evaluating and Recommending Connectors 11.3 Managing Approvals with Cloud Flows
12 Chapter 12: Cloud Flow Actions, Control, and Troubleshooting
12.1 Configuring Cloud Flow Actions 12.2 Implementing Flow Control with Conditions and Loops 12.3 Testing and Troubleshooting Cloud Flows
13 Chapter 13: Building Prompts in AI Hub
13.1 Building Prompts from Templates or from Blank 13.2 Adding Inputs to a Prompt 13.3 Adding Knowledge to a Prompt
14 Chapter 14: Prompt Settings, Models, and Consuming AI
14.1 Customizing Prompt Settings and Models 14.2 Consuming Prompts in Apps and Cloud Flows 14.3 Consuming AI Models in Apps and Cloud Flows
15 Chapter 15: Business Rules and Business Process Flows
15.1 Configuring Business Rules 15.2 Configuring Business Process Flows
16 Chapter 16: Calculated Columns and Business Logic Strategy
16.1 Calculated, Rollup, and Formula Columns 16.2 Evaluating Use Cases for Business Logic
13.1 Building Prompts from Templates or from Blank
Key Takeaways Explain This Topic AI
The AI hub in make.powerapps.com (and make.powerautomate.com) is the maker-portal home for prompts and AI models; a custom prompt starts at AI hub > Prompts > Build your own prompt
Every prompt has two parts: the instruction (what the model should do) and the context (the information it needs to do it)
Prompt builder prompts run on language models powered by Azure OpenAI Service, and availability can vary by region
The template gallery (summarize, classify, extract, reply, and more) preloads a working instruction, inputs, and sample test values that makers customize instead of starting from blank
A prompt becomes consumable from apps, flows, Copilot Studio topics, and Dataverse prompt columns only after it is saved
Last updated: July 2026
Prompts are how makers give natural-language instructions to a generative AI model. In the Power Platform, the AI hub is the maker-portal home for AI capabilities, and it hosts two distinct asset types the AB-410 exam expects you to tell apart: prompts , which are natural-language instructions run on language models powered by Azure OpenAI Service , and AI models , which are the prebuilt and custom AI Builder predictors such as invoice processing, sentiment analysis, and category classification.
Finding the AI Hub
Sign in to make.powerapps.com and select AI hub in the left navigation, then Prompts . The same hub exists in make.powerautomate.com , and the identical prompt designer appears inside Microsoft Copilot Studio when you add a prompt as a tool to an agent or topic. From the Prompts page you can either browse the template gallery or select Build your own prompt to open a blank designer.
Template gallery vs. blank prompt
The gallery offers starting points for the most common generative tasks:
- Template category What the model does Typical maker scenario
- Summarize Condenses long text into key points Roll up email threads or case notes
- Classify Assigns text to one of your defined categories Route inbound feedback by topic
- Extract Pulls structured fields out of unstructured text Get names, dates, or totals from free text
- Reply Drafts a response in a specified tone Suggest an answer to a customer email
Choosing a template preloads three things: a working instruction , ready-made inputs , and sample test values , so you can select Test immediately and then customize. Starting from blank gives you full control over the instruction, inputs, and knowledge — the right choice when no template matches the task, or when the output contract (format, length, tone) is strict.
The two parts of every prompt
Microsoft's prompt engineering guidance splits a prompt into the instruction and the context :
The instruction states what the model should do, for example: Summarize this email in three bullet points.
The context supplies what it needs to do it well, for example: The email contains customer feedback collected over the past week.
A custom prompt can answer questions, complete text, translate languages, summarize documents, or identify tasks and action items inside text. Complexity ranges from a single sentence to a detailed multi-paragraph specification.
Exam trap: do not confuse prompts with AI Builder models. Prebuilt and custom models are trained predictors managed under AI hub > AI models ; prompts are generative instructions authored under AI hub > Prompts . A question asking where to create a reusable GPT instruction points at Prompts — never at the Power Platform admin center , which governs environments, capacity, and tenant policies rather than authoring.
Prompt engineering the exam cares about
The exam tests whether you can recognize a well-formed prompt. Microsoft's guidance defines four characteristics: a good prompt is clear and concise , specific , contextual , and relevant . In practice that translates into concrete techniques you should be able to spot in scenario answers:
Instruction clarity — one task per prompt, unambiguous verbs. If a prompt both summarizes and drafts a reply, results degrade; split the work.
Structure and format directives — state the output contract: return exactly three sentences, respond as JSON with keys summary and sentiment, or answer in the same language as the input.
Few-shot examples — include one or more sample input/output pairs so the model mimics the pattern. This is the single most reliable way to stabilize formatting.
Tone and audience — name the reader and register, for example: write for a non-technical executive in a professional, reassuring tone.
Microsoft even publishes a downloadable prompt engineering guide for prompt builder, and the exam borrows its vocabulary — expect the terms instruction, context, and specific to appear in answer options.
Iterate with the Test button
The designer is built for a tight loop: enter a sample value for each input, select Test , read the response, adjust the instruction, and test again. Change one variable at a time so you know what fixed what. When generated text is too long or drifts off-topic, the documented fix is to make the prompt more specific — tighten the instruction and add explicit length or format directives rather than abandoning the prompt.
Naming, saving, and versioning at awareness level
Select the generated name at the top left of the designer to give the prompt a descriptive name — it becomes the identifier makers see in Power Automate actions and Copilot Studio tool pickers. Select Save to persist the prompt into the environment. At awareness level: re-saving overwrites the prompt in place, so teams typically manage iterations through naming conventions and by moving prompts between environments with solutions . Nothing is consumable until it has been saved.
Where saved prompts surface
Once saved, a prompt is a reusable AI asset across the platform:
Power Apps — call it from a canvas app through Power Fx.
Power Automate — add the AI Builder Run a prompt action inside a cloud flow.
Copilot Studio — add the prompt as a tool in a topic so an agent can invoke it.
Dataverse prompt columns — bind prompt logic to a table column whose value is generated from other columns.
Chapter 14 walks through those consumption paths in detail; for this chapter, anchor the authoring lifecycle: create in the AI hub, test iteratively, save, then consume. Also remember that prompt builder runs on Azure OpenAI-powered models, is limited to certain regions, and can be subject to usage limits and capacity throttling — fair game for awareness-level exam items.