Reusable Prompt Files Now Available in Visual Studio with GitHub Copilot

Microsoft has introduced reusable prompt files in Visual Studio, making it easier for developers to streamline their GitHub Copilot experience. This new feature allows teams to save, share, and reuse prompts, helping improve consistency and efficiency in coding workflows.

Here’s what you need to know:

• Why it matters
  Quickly access saved prompts at the prompt box, with automatic storage and sharing to help your team get more value from Copilot.

• Reusable prompt files simplify prompt-writing
  Write, save, and run prompts directly in Visual Studio by creating a .GitHub/prompts/[title].prompt.md file—no more digging through old notes or emails.

• Work faster and collaborate better
  Prompts are shared in your repository, ensuring consistency across projects and making them instantly available to your teammates.

• Tips for maximizing your reusable prompts
  Start with the awesome-copilot repo, use #prompt: or the context menu to add prompts, and organize them by topic or workflow. Regularly refine your prompts and encourage teammates to contribute for a shared library of best practices.

User-Defined Functions (UDFs) Are Now Generally Available in Power Apps

Microsoft has officially marked User-Defined Functions in Power Apps Canvas apps as Generally Available (GA), empowering developers to write, test, and maintain reusable logic components within their applications. This launch enhances app maintainability by allowing logic to be modularized into named functions, making larger Power Apps projects easier to manage and understand.

Here’s what you need to know:

• Why it matters:
  UDFs let you extract common logic into reusable functions with parameters. This reduces duplication, keeps logic in sync, and makes your apps easier to scale, maintain, and test.

• Example in action:

  • Create a simple UDF, like converting Fahrenheit to Celsius, and reuse it across multiple controls.

  • Add interactivity with sliders or labels to see results update instantly.

  • Go further by building UDFs that perform actions, like adding entries to a collection and displaying notifications.

Power BI Unveils "Semantic Model Refresh Templates" in Preview

Microsoft has announced a new preview feature in Power BI that streamlines how users manage and sequence semantic model refreshes. The update introduces built-in templates designed to make refresh workflows more intuitive and efficient.

With these templates, you can visualize and sequence refresh activities for both common and advanced scenarios:

• Access with ease

  Select Create advanced refresh from the refresh dropdown on your semantic model’s details page in your Microsoft Fabric workspace.

• Dataflow-driven refresh

  Trigger semantic model refreshes right after a dataflow refresh by configuring your workspace, connection, and model.

• Incremental refresh

  Refresh only updated partitions or tables, saving time and resources.

• Smart alerts

  Add follow-up actions like sending Microsoft Teams or Outlook messages after each execution.

• Flexible scheduling

  Automate refreshes with custom intervals: minute-by-minute, hourly, monthly, or even quarterly.

Microsoft Unveils September AI Roadmap for Visual Studio

Microsoft this week published its public roadmap for AI in Visual Studio, previewing the enhancements and agentic experiences developers can expect in September’s updates. The monthly roadmap aims to make Visual Studio “smarter, faster, and more intuitive,” while clarifying that the listed items reflect ongoing work—not firm commitments.

Here are some of the plans:

• New Modes for Debugging & Profiling – Easily switch between modes, with built-in and extension support designed to handle complex workflows.

• Customization via Modes – Seamlessly integrate your unique workflows across your stack, tailored to you and your organization.

• Model Context Protocol (MCP) – Bring your entire dev stack into Visual Studio with expanded capabilities for prompts, resources, and more—all backed by enterprise-grade security and governance.

• Models – Access the latest and most effective AI models, carefully evaluated to deliver the best experience in Visual Studio.

• Azure Agentic DevOps – Harness agentic workflows and Azure-native capabilities to build, deploy, and manage apps securely, with the governance you trust.

You can view the full list here.

Visual Studio August 2025 Update Brings AI and Debugging Upgrades

The latest Visual Studio 2025 update is here, delivering smarter AI features, improved debugging tools, and more control for developers. Microsoft’s August release focuses on boosting productivity and streamlining the developer experience.

Here’s what you need to know:

• GPT-5 Support

  Experience the latest in AI-powered development with GPT-5—delivering smarter, more accurate, and context-aware code suggestions and chat experiences.

• Model Context Protocol (MCP) GA

  Connect your entire stack seamlessly. MCP lets AI agents tap into databases, code search, and deployment systems—without building custom integrations.

• Smarter Copilot Chat

  Find the right code faster with improved semantic search, making Copilot Chat more reliable in surfacing relevant snippets.

• Google Sign-in for Copilot

  Getting started is easier than ever—sign up for Copilot directly with your Google account.

• Bring Your Own Model

  Use the LLM that works best for you. Connect your own models via API keys from OpenAI, Google, or Anthropic.

You can view the full list here.

GitHub Copilot Brings Next-Gen Diagnostics to .NET in Visual Studio

Microsoft has unveiled a powerful new diagnostics toolset for .NET development in Visual Studio, integrated with GitHub Copilot to streamline everything from breakpoint suggestions to LINQ query generation. The feature set—including context-aware tracepoint suggestions, immediate breakpoint troubleshooting, and a parallel stacks analysis & auto summaries—promises to significantly speed up debugging workflows and make profiling in Visual Studio more intuitive and efficient.

Here’s what you need to know:

• Breakpoint & Tracepoint Suggestions

  Skip manual setup — Copilot suggests the right conditional expressions or tracepoint actions based on your context.

• Breakpoint Troubleshooting

  No more trial-and-error. Copilot explains non-binding breakpoints, whether caused by mismatched symbols, build configs, or optimized code paths, and guides you to a fix.

• LINQ Query Hover

  Hover over any LINQ statement while debugging and get clear explanations, context evaluation, and insights on inefficiencies without leaving the editor.

• Parallel Stacks Analysis & Auto Summaries

  Untangle async and multithreaded code with Copilot’s thread summaries, app state insights, and deadlock/hang detection.

• Auto Insights in CPU & Allocation Tools

  Copilot highlights hot paths, high-usage functions, and performance bottlenecks so you can zero in on what really matters.

You can view the full list here.

Microsoft Brings Azure Tools into Copilot Agent Mode via Azure MCP in VS 2022

Microsoft has rolled out the public preview of GitHub Copilot for Azure in Visual Studio 2022 (version 17.14+), integrating a suite of Azure developer tools directly into Copilot's Agent Mode. The extension streamlines developer workflows by automatically installing and managing an Azure MCP server, enabling users to query resources, run CLI commands, deploy via azd, and diagnose issues—all within the Copilot Chat.

Here’s what you need to know:

• What’s in Public Preview?

  You can now use a zero-setup Azure MCP server that starts automatically when you open Copilot Chat, agent mode with Azure tools that pick the right resources for your goals, and broad Azure coverage via MCP tools.

• Getting Started

  Make sure you’re on Visual Studio 2022 (v17.4 or later) with an active GitHub Copilot subscription, Copilot Chat enabled, and a Microsoft account linked to an Azure subscription.

• How to Install

  • Step 1: Install the GitHub Copilot for Azure (Preview) extension in Visual Studio—the Azure MCP server starts automatically, so no manual setup is needed.

  • Step 2: Open Copilot Chat and select Agent Mode.

  • Step 3: Enable the Azure Extension under Select Tools and include resource details like subscription, resource group, and name in your prompts for the best results.

Model Context Protocol Support Now Generally Available in Visual Studio

Model Context Protocol (MCP) support is now generally available in Visual Studio, unlocking rich, real-time integration with external tools and services to power smarter AI-driven workflows. Now enabled by default in Copilot Chat, MCP lets developers connect securely to databases, deployment pipelines, and custom servers—transforming Visual Studio into a powerful hub for intelligent development.

Here’s what you need to know:

• What is MCP?
  MCP is an open standard that allows AI agents to seamlessly connect with external tools and services—databases, code search, deployment pipelines, and more—through a consistent, secure protocol.

• Why it matters
  It enables custom workflows, enterprise-grade integrations, and smarter, context-aware conversations.

• Full MCP client integration
  Connect to local or remote MCP servers via .mcp.json. Visual Studio now makes managing server inputs, authentication, and tools a first-class experience.

• First-class authentication
  Securely sign in to protected endpoints with built-in Visual Studio flows. Support extends across the full MCP authentication spec—no matter the provider.

• One-click server installation
  Add new MCP servers straight from the web in just a click—making it easier than ever to expand your AI-powered workflows.

Take Control Over Your Copilot Edit Suggestions in Visual Studio

Visual Studio has just rolled out enhanced settings that let developers fine-tune when and how GitHub Copilot code suggestions appear—whether that's pausing suggestions while typing or triggering them only with a keystroke. This update aims to strike a better balance between helpful AI assistance and keeping your coding flow uninterrupted by unwanted distractions.

Here’s what you need to know:

• Pause suggestions while typing

  Keep your flow uninterrupted. Suggestions will only appear after you pause, so you can type freely without constant pop-ups. Simply go to Tools -> Options → IntelliCode → Advanced and turn on wait for pauses in typing before showing whole line completions.

• Request completions only when you’re ready

  Disable automatic completions and trigger them manually with shortcuts like Alt + , or Alt + .. Perfect for developers who prefer fewer distractions.

• Preview edit suggestions on demand

  No more auto-inserts. Review the next edit suggestion only when you click the margin indicator or press Tab. Accept with Tab or dismiss with ESC—you’re in charge.

• Accept code in smaller pieces

  Take partial control of suggestions. Accept one word at a time with Ctrl + → or one line at a time with Ctrl + ↓, tailoring the output to your needs.

Visual Studio Expands Chat with Custom AI Model Integration

Microsoft has announced that Visual Studio Chat now supports third-party language models via API keys from providers like OpenAI, Anthropic, and Google—offering developers enhanced choice, customization, and control. This marks a significant shift, letting users seamlessly switch between built-in Copilot models and their own preferred AI models right inside the IDE.

Here’s what you need to know:

• Why it matters

  • Try the latest models as soon as they’re released

  • Customize models to fit your security, infrastructure, or performance needs

  • Manage usage, quotas, and billing directly with your provider

  • Switch seamlessly between Copilot’s built-in models and your own

• How to set it up

  1. Open the Chat Window in Visual Studio.

  2. From the model picker, select Manage models.

  3. Pick your provider and enter your API key.

  4. Choose from the preset options or add a custom model name.