Power Platform AI Prompts Quick Overview
What is a Prompt?
A prompt is a natural language instruction that tells a large language model (LLM) to perform a task. The model follows the prompt to determine the structure and content of the text it needs to generate.
Here is the link to Microsoft documentation.
Power Platform Prebuilt Prompt
Some prebuilt prompts are available on the Power Apps Maker portal :
- Respond to a complaint
- Classify text
- Extract information from text
- Summarize text
It is also possible create custom prompts. AI Builder provides a prompt builder that allow makers to write, test, save, and use prompts to generate text.

Example with "Respond to a complaint" prompt

Please note that human supervision is a crucial step in handling content generated by a GPT model. Models like GPT are trained on vast datasets, but AI-generated content may still contain misleading or fabricated information and are susceptible to prompt injection attacks. Therefore, it's essential for a human to review it before posting online, sharing with customers, or making business decisions based on it. This oversight not only helps detect potential errors and biases but also ensures the content aligns with the intended use case and reflects the company's values.
Furthermore, human review can pinpoint issues with the GPT model itself. For instance, if the model produces irrelevant content, adjusting the prompt may be necessary.
Where to use prompts?
As of the time of writing, AI Prompts can be used in Power Apps (canvas) or in Power Automate.
Distribute prompts
Prompts can be distributed across environments using solutions. Prompts can be added using menu Add existing > AI models inside a solution.
Published on:
Learn moreWe can help you with Power Platform AI Prompts Quick Overview
If you want help implementing, troubleshooting, or improving this product, contact us and we’ll point you in the right direction.
Related posts
Unmanaged developer environments are ending: what to do before December 15
Unmanaged developer environments can no longer be created, are suspended on December 15, 2026 and deleted in January. Who is affected and how ...
Backfill Dataverse prompt columns on existing records
Prompt columns only ran on new or updated rows. A bulk option now fills existing records as a system job: the two settings, the steps and the ...
Reveal a masked Dataverse column with Dataverse.Unmask
A new Power Fx function lets authorised users of a canvas app reveal one masked Dataverse value on demand, while column security stays in char...
Generative pages in model-driven apps: a practical guide
What generative pages are, where they stand in November 2026, the two ways to build them, the code they produce, and the limits to plan for.
Big Update — CRMJetty Dynamics 365 Portal Now Comes with Powerful New Integrations
Enterprise portals are evolving rapidly. They are no longer mere portals through which CRM records can be viewed. Portals are now expected to ...
Power Apps Code Apps in Practice: Recreating Model-Driven App Patterns
Power Apps Code Apps went from preview to general availability in record time. With that, Microsoft has sent a clear signal that it is taking ...
Use XRM type definition in your Javascript file
We implement Javascript webresource for form scripting (form onload, field on change,...). Our code need to follow the official document Xrm o...
Dataverse Functions Deep Dive
Microsoft released in 2024 Dataverse low-code plugins. The promise was great : writing plugins using low-code tools. No need to have developme...