AGENT SKILL

winmd-api-search

Find and explore Windows desktop APIs. Use when building features that need platform capabilities — camera, file access, notifications, UI controls, AI/ML, sensors, networking, etc. Discovers the right API for a task and retrieves full type details (methods, properties, events, enumeration values).

适用职业开发工程师
使用场景后端与接口
License 许可
MIT
所属仓库
TorchiaHub/Skills
最近更新
2026-06-21
Stars
★ 0
Forks
0

01 前言 Front Matter

name
winmd-api-search
license
Complete terms in LICENSE.txt
description
Find and explore Windows desktop APIs. Use when building features that need platform capabilities — camera, file access, notifications, UI controls, AI/ML, sensors, networking, etc. Discovers the right API for a task and retrieves full type details (methods, properties, events, enumeration values).

02 资源文件 8

03 SKILL.md 正文

WinMD API Search

This skill helps you find the right Windows API for any capability and get its full details. It searches a local cache of all WinMD metadata from:

  • Windows Platform SDK — all Windows.* WinRT APIs (always available, no restore needed)
  • WinAppSDK / WinUI — bundled as a baseline in the cache generator (always available, no restore needed)
  • NuGet packages — any additional packages in restored projects that contain .winmd files
  • Project-output WinMD — class libraries (C++/WinRT, C#) that produce .winmd as build output

Even on a fresh clone with no restore or build, you still get full Platform SDK + WinAppSDK coverage.

When to Use This Skill

  • User wants to build a feature and you need to find which API provides that capability
  • User asks "how do I do X?" where X involves a platform feature (camera, files, notifications, sensors, AI, etc.)
  • You need the exact methods, properties, events, or enumeration values of a type before writing code
  • You're unsure which control, class, or interface to use for a UI or system task

Prerequisites

  • .NET SDK 8.0 or later — required to build the cache generator. Install from dotnet.microsoft.com if not available.

Cache Setup (Required Before First Use)

All query and search commands read from a local JSON cache. You must generate the cache before running any queries.

# All projects in the repo (recommended for first run)
.\.github\skills\winmd-api-search\scripts\Update-WinMdCache.ps1

# Single project
.\.github\skills\winmd-api-search\scripts\Update-WinMdCache.ps1 -ProjectDir <project-folder>

No project restore or build is needed for baseline coverage (Platform SDK + WinAppSDK). For additional NuGet packages, the project needs dotnet restore (which generates project.assets.json) or a packages.config file.

Cache is stored at Generated Files\winmd-cache\, deduplicated per-package+version.

What gets indexed

Source When available
Windows Platform SDK Always (reads from local SDK install)
WinAppSDK (latest) Always (bundled as baseline in cache generator)
WinAppSDK Runtime When installed on the system (detected via Get-AppxPackage)
Project NuGet packages After dotnet restore or with packages.config
Project-output .winmd After project build (class libraries that produce WinMD)

Note: This cache directory should be in .gitignore — it's generated, not source.

How to Use

Pick the path that matches the situation:


Discover — "I don't know which API to use"

The user describes a capability in their own words. You need to find the right API.

0. Ensure the cache exists

If the cache hasn't been generated yet, run Update-WinMdCache.ps1 first — see Cache Setup above.

1. Translate user language → search keywords

Map the user's daily language to programming terms. Try multiple variations:

User says Search keywords to try (in order)
"take a picture" camera, capture, photo, MediaCapture
"load from disk" file open, picker, FileOpen, StorageFile
"describe what's in it" image description, Vision, Recognition
"show a popup" dialog, flyout, popup, ContentDialog
"drag and drop" drag, drop, DragDrop
"save settings" settings, ApplicationData, LocalSettings

Start with simple everyday words. If results are weak or irrelevant, try the more technical variation.

2. Run searches

.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action search -Query "<keyword>"

This returns ranked namespaces with top matching types and the JSON file path.

If results have low scores (below 60) or are irrelevant, fall back to searching online documentation:

  1. Use web search to find the right API on Microsoft Learn, for example:
    • site:learn.microsoft.com/uwp/api <capability keywords> for Windows.* APIs
    • site:learn.microsoft.com/windows/windows-app-sdk/api/winrt <capability keywords> for Microsoft.* WinAppSDK APIs
  2. Read the documentation pages to identify which type matches the user's requirement.
  3. Once you know the type name, come back and use -Action members or -Action enums to get the exact local signatures.

3. Read the JSON to choose the right API

Read the file at the path(s) from the top results. The JSON has all types in that namespace — full members, signatures, parameters, return types, enumeration values.

Read and decide which types and members fit the user's requirement.

4. Look up official documentation for context

The cache contains only signatures — no descriptions or usage guidance. For explanations, examples, and remarks, look up the type on Microsoft Learn:

Namespace prefix Documentation base URL
Windows.* https://learn.microsoft.com/uwp/api/{fully.qualified.typename}
Microsoft.* (WinAppSDK) https://learn.microsoft.com/windows/windows-app-sdk/api/winrt/{fully.qualified.typename}

For example, Microsoft.UI.Xaml.Controls.NavigationView maps to: https://learn.microsoft.com/windows/windows-app-sdk/api/winrt/microsoft.ui.xaml.controls.navigationview

5. Use the API knowledge to answer or write code


Lookup — "I know the API, show me the details"

You already know (or suspect) the type or namespace name. Go direct:

# Get all members of a known type
.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action members -TypeName "Microsoft.UI.Xaml.Controls.NavigationView"

# Get enum values
.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action enums -TypeName "Microsoft.UI.Xaml.Visibility"

# List all types in a namespace
.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action types -Namespace "Microsoft.UI.Xaml.Controls"

# Browse namespaces
.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action namespaces -Filter "Microsoft.UI"

If you need full detail beyond what -Action members shows, use -Action search to get the JSON file path, then read the JSON file directly.


Other Commands

# List cached projects
.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action projects

# List packages for a project
.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action packages

# Show stats
.\.github\skills\winmd-api-search\scripts\Invoke-WinMdQuery.ps1 -Action stats

If only one project is cached, -Project is auto-selected. If multiple projects exist, add -Project <name> (use -Action projects to see available names). In scan mode, manifest names include a short hash suffix to avoid collisions; you can pass the base project name without the suffix if it's unambiguous.

Search Scoring

The search ranks type names and member names against your query:

Score Match type Example
100 Exact name Button → Button
80 Starts with Navigation → NavigationView
60 Contains Dialog → ContentDialog
50 PascalCase initials ASB → AutoSuggestBox
40 Multi-keyword AND navigation item → NavigationViewItem
20 Fuzzy character match NavVw → NavigationView

Results are grouped by namespace. Higher-scored namespaces appear first.

Troubleshooting

Issue Fix
"Cache not found" Run Update-WinMdCache.ps1
"Multiple projects cached" Add -Project <name>
"Namespace not found" Use -Action namespaces to list available ones
"Type not found" Use fully qualified name (e.g., Microsoft.UI.Xaml.Controls.Button)
Stale after NuGet update Re-run Update-WinMdCache.ps1
Cache in git history Add Generated Files/ to .gitignore

References

04 同仓库更多 Skills

TorchiaHub
TorchiaHub/Skills

resemble-detect

Deepfake detection and media safety — detect AI-generated audio, images, video, and text, trace synthesis sources, apply watermarks, verify speaker identity, and analyze media intelligence using Resemble AI

后端与接口图像与多媒体
★ 0
TorchiaHub
TorchiaHub/Skills

refactor

Surgical code refactoring to improve maintainability without changing behavior. Covers extracting functions, renaming variables, breaking down god functions, improving type safety, eliminating code smells, and applying design patterns. Less drastic than repo-rebuilder; use for gradual improvements.

代码审查视觉设计
★ 0
TorchiaHub
TorchiaHub/Skills

quality-playbook

Explore any codebase from scratch and generate six quality artifacts: a quality constitution (QUALITY.md), spec-traced functional tests, a code review protocol with regression test generation, an integration testing protocol, a multi-model spec audit (Council of Three), and an AI bootstrap file (AGENTS.md). Includes state machine completeness analysis and missing safeguard detection. Works with any language (Python, Java, Scala, TypeScript, Go, Rust, etc.). Use this skill whenever the user asks to set up a quality playbook, generate functional tests from specifications, create a quality constitution, build testing protocols, audit code against specs, or establish a repeatable quality system for a project. Also trigger when the user mentions 'quality playbook', 'spec audit', 'Council of Three', 'fitness-to-purpose', 'coverage theater', or wants to go beyond basic test generation to build a full quality system grounded in their actual codebase.

代码审查测试与质量
★ 0
TorchiaHub
TorchiaHub/Skills

prd

Generate high-quality Product Requirements Documents (PRDs) for software systems and AI-powered features. Includes executive summaries, user stories, technical specifications, and risk analysis.

产品经理
★ 0
TorchiaHub
TorchiaHub/Skills

power-platform-architect

Use this skill when the user needs to transform business requirements, use case descriptions, or meeting transcripts into a technical Power Platform solution architecture, including component selection and Mermaid.js diagrams.

★ 0
TorchiaHub
TorchiaHub/Skills

plantuml-ascii

Generate ASCII art diagrams using PlantUML text mode. Use when user asks to create ASCII diagrams, text-based diagrams, terminal-friendly diagrams, or mentions plantuml ascii, text diagram, ascii art diagram. Supports: Converting PlantUML diagrams to ASCII art, Creating sequence diagrams, class diagrams, flowcharts in ASCII format, Generating Unicode-enhanced ASCII art with -utxt flag

★ 0