What problem does it solve? Building a high-quality agent skill for a Microsoft technology requires deep, accurate knowledge of that technology's concepts, APIs, and best practices, which is time-consuming to research and structure manually. ## Core Features & Use Cases - Guided Investigation Workflow: Uses Microsoft Learn MCP tools (docs search, docs fetch, code sample search) in a three-phase research process covering scope discovery, core content, and depth topics. - Ready-to-Use Templates: Provides four skill templates for SDKs/libraries, Azure services, frameworks/platforms, and APIs/protocols, with a selection guide. - Hybrid Content Strategy: Balances locally stored foundational knowledge with dynamic Learn MCP lookups so generated skills stay concise yet complete. - CLI Fallback: Supports the mslearn CLI (npx @microsoft/learn-cli) when the Learn MCP server is unavailable. - Use Case: Ask the agent to create a skill for Semantic Kernel; it researches official docs, gathers C# code samples, and generates a SKILL.md with key concepts, quick-start code, and search queries for deeper topics. ## Quick Start Ask the agent to create a skill for a Microsoft technology such as Azure Cosmos DB, and it will investigate the official documentation and generate a complete skill directory.
