Microsoft Foundry

Home » Microsoft Foundry

GitHub Agentic Workflows: The Next Evolution of Repository Automation for Power Platform and Enterprise Developers

GitHub Agentic Workflows automating repository management and Power Platform development — AI agents handling issue triage, documentation, testing, and deployment

Traditional GitHub Actions workflows are powerful but deterministic — they can only do what you explicitly tell them. GitHub Agentic Workflows change that. Launched in February 2026, they bring AI coding agents (Copilot CLI, Claude Code, Codex) into GitHub Actions with natural language specifications, security guardrails, and human review. Instead of writing imperative YAML, you describe outcomes in plain language and let AI agents figure out the implementation. This guide explores the six core automation categories and practical applications for Power Platform developers managing PCF controls, Dataverse solutions, and ALM pipelines.

GitHub Copilot Workspace Context: The Complete Developer Guide to Smarter AI Coding

GitHub Copilot Workspace Context: The Complete Developer Guide to Smarter AI Coding

GitHub Copilot is only as smart as the context it can see. This guide explains exactly how workspace context and codebase indexing work in VS Code — remote vs local vs basic indexes, when to use #codebase vs @workspace, how to write copilot-instructions.md for your team, and six prompt patterns that consistently produce better answers. Practical, code-first, no fluff.

The Risk Decision Matrix for GitHub Copilot CLI Autopilot Mode in Power Platform

The Risk Decision Matrix for GitHub Copilot CLI Autopilot Mode in Power Platform

GitHub Copilot CLI Autopilot mode reached GA in February 2026 — but for Power Platform, autonomous PAC CLI execution carries real risk. This article builds a scored Risk Decision Matrix mapping every PAC CLI operation to one of four zones: full Autopilot, conditional, Plan Mode only, or human-only. Includes a calculable Autopilot Safety Score formula, a ready-to-use AGENTS.md template, and preToolUse hook enforcement patterns grounded in Singapore IMDA, AWS, and McKinsey agentic governance frameworks.

Claude Skills: The Complete Developer Guide to Building Reusable AI Workflows (With Code)

Claude Skills: The Complete Developer Guide to Building Reusable AI Workflows (With Code)

Claude Skills are reusable, Markdown-based instruction bundles that turn Claude into a specialist for any domain — automatically loaded when relevant, no re-prompting required. In this complete developer guide, we break down all 17 available skills across document automation, frontend design, generative art, MCP server building, and more. Each section includes working Python and TypeScript code samples, dependency lists, and real integration patterns so you can drop them into your own projects today. Whether you're generating Word docs with docx, building MCP servers with mcp-builder, or creating animated Slack GIFs with slack-gif-creator, this guide shows you exactly how each skill works under the hood — and how to build your own.

GitHub Copilot CLI Is Now Generally Available — And Power Platform Developers Should Pay Attention

GitHub Copilot CLI Is Now Generally Available — And Power Platform Developers Should Pay Attention

GitHub Copilot CLI is now generally available for all paid Copilot subscribers. This article breaks down what that means for Power Platform developers — covering the official Microsoft Power Platform Skills plugin, real-world PAC CLI and PCF scenarios, Agent Skills ecosystem by developer profile, and a full Copilot CLI vs Claude Code comparison.

Stop Writing Vague AI Prompts: The Developer’s Playbook for Mastering agents.md in 2026

agents.md best practices for GitHub Copilot specialist agents

Your AI agent just modified the wrong file. Again. Before you blame the model — when did you last check what instructions you actually gave it?
Most developers hand a powerful coding agent a vague persona and wonder why results are inconsistent. GitHub's analysis of 2,500+ repositories reveals the uncomfortable truth: the gap between agents that work and agents that guess comes down to one file you're probably not writing correctly.
In this guide: the six pillars of a high-performing agents.md, a production-ready Dataverse plugin agent template, and the maintenance habit that keeps your AI sharp as your codebase grows. Your agent is only as good as the manual you write for it.

GitHub Copilot Agentic Memory: What It Is, How It Works?

GitHub Copilot Agentic Memory vs Stateless AI

GitHub Copilot just launched a persistent, repository-scoped agentic memory system in public preview. Instead of starting from scratch every session, Copilot now learns your codebase conventions over time and shares that knowledge across the coding agent, code review, and CLI. In this article, we explore how it works under the hood — and walk through three concrete Power Platform scenarios (PCF control development, Power Automate solution management, and Dataverse architecture patterns) where this changes the daily developer experience significantly.

Power Apps MCP Server: Complete Guide to Every Feature (Public Preview 2026)

Power Apps MCP Server agent feed — AI agent extracting Dataverse records with human approval interface — AIDevMe 2026

The Power Apps MCP (Model Context Protocol) Server is now in public preview. It exposes three tools — invoke_data_entry, request_assistance, and log_for_review — that let AI agents automate tasks inside model-driven apps with built-in human supervision via a redesigned agent feed. This guide covers every feature, technical detail, current limitation, and step-by-step setup path.

Terraform for Power Platform Developers: Provision Azure Resources & Automate CI/CD

Terraform for Power Platform Developers: Provision Azure Resources & Automate CI/CD

This guide shows Power Platform developers how to use Terraform to provision and manage Azure resources — Key Vault, Storage, Function Apps, API Management, and App Registrations — and integrate the full infrastructure lifecycle into GitHub Actions and Azure DevOps pipelines. All code samples are production-ready and environment-parameterized.