skill-creator

Guide for creating effective GitHub Copilot skills (.github/skills/) in this repository. Use when creating a new skill, updating an existing skill, or when…

npx skills add https://github.com/microsoft/vscode-documentdb --skill skill-creator

Skill Creator for GitHub Copilot

Create and maintain skills in .github/skills/ that extend GitHub Copilot's capabilities with project-specific knowledge.

What Are Skills

Skills are SKILL.md files that provide specialized, procedural knowledge that Copilot doesn't inherently have. They turn Copilot from a general assistant into a domain expert for specific tasks in this codebase.

Skills provide:

  • Specialized workflows — multi-step procedures for project-specific domains
  • Domain expertise — architecture patterns, conventions, business logic
  • Bundled references — detailed docs loaded only when needed

Skill Anatomy

.github/skills/
└── skill-name/
    ├── SKILL.md              (required — frontmatter + instructions)
    └── references/           (optional — detailed docs, loaded on demand)

SKILL.md Structure

---
name: my-skill-name
description: What this skill does and WHEN to use it. Include trigger words and scenarios. This is the primary mechanism for Copilot to decide whether to load the skill body.
---

# Skill Title

Concise instructions for using this skill.

## When to Use

- Bullet list of triggering scenarios

## Core Content

(workflow, patterns, examples)

Wiring the Skill

After creating the SKILL.md, register it in .github/copilot-instructions.md under the <skills> section:

<skill>
<name>my-skill-name</name>
<description>Same or similar description as in SKILL.md frontmatter</description>
<file>${workspaceFolder}\.github\skills\my-skill-name\SKILL.md</file>
</skill>

Note: Replace ${workspaceFolder} with the absolute path to the repository root on your machine (e.g. \home\user\repos\vscode-documentdb). VS Code resolves skill file paths at runtime and currently requires absolute paths.

Copilot reads skill metadata (name + description) to decide when to trigger. The SKILL.md body is loaded only after triggering.

Core Principles

1. Concise is Key

The context window is shared with conversation history, other instructions, and user requests. Challenge each paragraph: "Does Copilot already know this?" and "Does this justify its token cost?"

  • Prefer concise examples over verbose explanations
  • Only include what Copilot doesn't already know — skip general TypeScript/React knowledge
  • Target under 300 lines for SKILL.md body; split into references if larger

2. Progressive Disclosure

Use a three-level loading system:

  1. Metadata (name + description) — always visible (~50 words)
  2. SKILL.md body — loaded when skill triggers
  3. References — loaded on demand by Copilot when deeper detail is needed

For skills with multiple variants or extensive reference material, keep the core workflow in SKILL.md and move details to references/:

## Advanced Topics

- **Detailed format spec**: See [FORMAT.md](./FORMAT.md)
- **Migration patterns**: See [references/migration.md](./references/migration.md)

3. Match Freedom to Task Fragility

Freedom LevelWhenExample
High (text guidance)Multiple valid approachesArchitecture recommendations
Medium (patterns with examples)Preferred pattern existstRPC router creation
Low (exact steps)Fragile, error-proneRelease note formatting

4. Do NOT Include

  • README.md, CHANGELOG.md, or auxiliary documentation
  • Setup/installation instructions for the skill itself
  • User-facing documentation — skills are for Copilot, not humans
  • Information Copilot already knows (general language features, common libraries)

Skill Creation Workflow

Step 1: Understand the Domain

Identify concrete scenarios the skill addresses:

  • What tasks trigger it?
  • What does Copilot get wrong without it?
  • What project-specific knowledge is needed?

Step 2: Plan Contents

For each scenario, determine:

  • What code patterns or templates are repeatedly needed?
  • What reference material should be available?
  • What pitfalls must be called out?

Step 3: Write the Skill

  1. Create .github/skills/{skill-name}/SKILL.md
  2. Write frontmatter: name and description (description is the trigger — be comprehensive about when to use)
  3. Write body: Core workflow, patterns, examples. Keep it lean.
  4. Add references (optional): Split out detailed format specs, schemas, or variant-specific docs
  5. Register in .github/copilot-instructions.md

Step 4: Validate

  • Ensure the description covers all trigger scenarios
  • Verify code examples follow project conventions (see .github/copilot-instructions.md and .github/instructions/)
  • Check that referenced files exist and paths are correct
  • Confirm the skill doesn't duplicate information already in .github/instructions/ files

Existing Skills Reference

SkillPurpose
writing-release-notesRelease notes and changelog generation
accessibility-aria-expertAccessibility issues in React/Fluent UI webviews
webview-trpc-messagingtRPC communication between extension host and webviews

Example: Minimal Skill

---
name: my-pattern
description: Implements the XYZ pattern for this codebase. Use when creating new XYZ instances, modifying existing XYZ behavior, or debugging XYZ-related issues.
---

# XYZ Pattern

## When to Use

- Creating a new XYZ
- Modifying XYZ behavior
- Debugging XYZ issues

## Pattern

\`\`\`typescript
// core pattern example
\`\`\`

## Common Pitfalls

- Don't do X because Y
- Always do Z when W

More skills from microsoft

oss-growth
microsoft
OSS growth hacker persona
agent-framework-azure-ai-py
microsoft
Build Azure AI Foundry agents using the Microsoft Agent Framework Python SDK (agent-framework-azure-ai). Use when creating persistent agents with AzureAIAgentsProvider, using hosted tools (code interpreter, file search, web search), integrating MCP servers, managing conversation threads, or implementing streaming responses. Covers function tools, structured outputs, and multi-tool agents.
development
airunway-aks-setup
microsoft
Set up AI Runway on AKS — from bare cluster to running model. Covers cluster verification, controller install, GPU assessment, provider setup, and first deployment. WHEN: "setup AI Runway", "onboard AKS cluster", "install AI Runway", "airunway setup", "deploy model to AKS", "GPU inference on AKS", "KAITO setup on AKS", "run LLM on AKS", "vLLM on AKS", "set up model serving on AKS", "AI Runway controller".
devops
appinsights-instrumentation
microsoft
Guidance for instrumenting webapps with Azure Application Insights. Provides telemetry patterns, SDK setup, and configuration references. WHEN: how to instrument app, App Insights SDK, telemetry patterns, what is App Insights, Application Insights guidance, instrumentation examples, APM best practices.
devops
applicationinsights-web-ts
microsoft
Instrument browser/web apps with the Application Insights JavaScript SDK (@microsoft/applicationinsights-web). Use for Real User Monitoring (RUM) — page views, clicks, AJAX/fetch dependencies, exceptions, custom events, and browser-side GenAI agent traces correlated to backend OpenTelemetry traces. Covers SDK Loader Script and npm setup, framework extensions (React, React Native, Angular), Click Analytics, telemetry initializers, and OTel GenAI semantic conventions for agent/tool/model spans emitted from the browser.
devops
azure-ai-anomalydetector-java
microsoft
Build anomaly detection applications with Azure AI Anomaly Detector SDK for Java. Use when implementing univariate/multivariate anomaly detection, time-series analysis, or AI-powered monitoring.
development
azure-ai-language-conversations-py
microsoft
Implement Conversational Language Understanding (CLU) using the azure-ai-language-conversations Python SDK. Use when working with ConversationAnalysisClient to analyze conversation intent and entities, building NLP features, or integrating language understanding into applications.
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python. Use for ML workspaces, jobs, models, datasets, compute, and pipelines. Triggers: "azure-ai-ml", "MLClient", "workspace", "model registry", "training jobs", "datasets".
development