AgentSkillsCN

module-author

为 CopilotTraining 研讨会模块生成包含练习规划表格的模块 README.md 文件。通过调研相关 URL 了解新功能特性,并据此制定战略性内容大纲。利用“练习作者”技能来具体实施各项练习。

SKILL.md
--- frontmatter
name: module-author
description: Generates module README.md with exercise planning table for CopilotTraining workshop modules. Researches URLs to understand new capabilities and creates strategic content outline. Use exercise-author skill to implement individual exercises.

Module Author Skill

Generate module README.md with exercise planning for CopilotTraining workshop modules. This skill focuses on module strategy and planning—the exercise-author skill handles tactical implementation.


When to Use This Skill

  • Creating a new training module from scratch
  • Updating module content when GitHub releases new capabilities
  • Researching and planning exercises for a specific feature

Invocation examples:


Core Module Design Principles

Three Essential Goals

Every module must deliver on these three goals:

  1. Teach one capability — Each module = one GitHub Copilot feature or technique
  2. Hands-on practice — At least one exercise (typically 2-3) to let readers see/touch/use it
  3. Persona relevance — Show why this matters to each persona type in a compelling way

Single-File Structure (README.md)

Each module README.md (~100-150 lines) contains:

  1. Story So Far — Progression from previous modules
  2. Why This Matters — Relevance to ALL personas (Sarah, David, Marcus, Elena, Rafael)
  3. What You'll Learn — Brief capability overview
  4. 📋 Exercise Planning Table — Detailed exercise specifications
  5. What This Feature Does — Technical overview
  6. Next Up — Preview of next module

Exercise files are generated separately by the exercise-author skill using the planning table as input.


Core Principles (from copilot-instructions.md)

Before generating, review .github/copilot-instructions.md for:

  • Training philosophy (Syntax Wizards → Markdown Whisperers)
  • Persona characteristics and voice guidelines
  • Metrics requirements (quantifiable improvements)
  • Official docs requirements (1-3 GitHub/MS links)
  • Emoji vocabulary for consistent visual language
  • Section separators (--- between all major sections)
  • Outcome-based language (professionals demonstrating results, not students learning)

URL Research Workflow

When URL is Provided

If the user provides a URL (blog post, docs page, announcement):

  1. Extract key information:

    • What capability is being introduced/updated?
    • What problem does it solve?
    • What are the main features or commands?
    • Any code examples or workflows shown?
    • Official documentation links referenced?
  2. Identify persona relevance:

    • How would Sarah validate this with metrics?
    • How would David use his expertise with this?
    • What workflow friction does this remove for Marcus?
    • How does this affect Elena's testing process?
    • What business value does Rafael see?
  3. Plan exercises:

    • What hands-on demonstrations would prove the value?
    • What artifacts would learners create?
    • What before/after metrics would be compelling?
    • Which persona should lead each exercise?
  4. Find official docs:

    • Primary documentation URL
    • Related API/CLI reference
    • Best practices guide

Template Structure

Use TEMPLATE.md as the base structure. The template includes all necessary sections with placeholder text and formatting.

Key Sections to Fill

1. Story So Far (2-3 paragraphs)

  • Summarize what was built in 2-3 previous modules
  • Explain how those capabilities are currently being used
  • Introduce the new challenge this module addresses
  • Add Integration Note connecting to specific artifacts from previous module

2. Why This Matters (All 5 Personas)

  • Write one outcome-focused statement per persona
  • Include specific metrics or measurable improvements
  • Show how capability delivers value to each role
  • See persona voice guidelines in .github/copilot-instructions.md

3. Exercise Planning Table (2-3 exercises)

  • One row per exercise with all columns filled
  • Lead persona marked with ⭐, support personas with 🤝
  • Problem column: current pain with specific metrics
  • Solution column: brief description of what will be built
  • Key Metrics column: before→after improvements
  • Artifacts column: specific file paths that will be created

4. What This Feature Does (Technical Overview)

  • 1-2 sentence capability explanation
  • Specific use case ("when questions span X files...")
  • 3 concrete artifacts learners will build
  • 2-3 official documentation links
  • Optional callout for critical tools/commands

5. Next Module Preview

  • One sentence describing next capability
  • Quote from a persona teasing the next challenge

Exercise Table Column Specifications

When populating the exercise planning table in TEMPLATE.md:

# (Exercise Number):

  • Format: [N.X](exercise-0X.md) where N = module number
  • Links to exercise file (created by exercise-author skill)

Exercise (Title):

  • Action-oriented, specific outcome
  • Example: "Enforcing Review Standards" not "Learn Custom Instructions"

Lead:

  • Primary persona (one only) with ⭐
  • Whose story anchors this exercise

Support:

  • 1-3 supporting personas with 🤝
  • Who collaborates or benefits

Problem:

  • Current pain point with specific metrics
  • Format: "[Issue]: [time], [count], [rounds/cycles]"
  • Example: "Manual reviews: 45 min/review, 8 violations/PR, 3 rounds"

Solution:

  • Brief description of what will be built
  • What capability solves the problem
  • Example: "Custom instructions for React standards"

Key Metrics:

  • Target improvements in shorthand
  • Format: "before→after" for each metric
  • Example: "45→5 min, 8→0 violations, 3→1 rounds"

Exercise Planning Guidelines

Step Count Goal: 3 Steps per Exercise

Each exercise specification should assume ~3 steps in implementation:

  • Ideal: 3 substantial, clear steps
  • Acceptable: 2 steps if naturally cohesive, or 4-5 if complexity demands it
  • Consider splitting: If you plan for more than 5 steps, consider if this should be 2 exercises

Persona Distribution Guidelines

Balance Across Module

Aim for diverse representation:

  • Each persona should lead at least 1 exercise across every 2-3 modules
  • Sarah and David typically lead architecture/standards exercises
  • Marcus leads infrastructure/workflow exercises
  • Elena leads testing/quality exercises
  • Rafael leads requirements/analysis exercises

Supporting Cast Selection

Choose support personas based on realistic collaboration:

  • Sarah + David — Architecture and standards alignment
  • Marcus + Elena — Testing infrastructure and automation
  • Sarah + Marcus — Standards enforcement in CI/CD
  • Elena + Sarah — Quality standards and review processes
  • Rafael + [any] — Business requirements meet technical implementation

Module Continuity Patterns

Story So Far

  • Summarize 2-3 previous modules in brief paragraphs
  • Explain what capabilities were added and how they're being used
  • Introduce the new challenge this module solves
  • Add Integration Note showing how this builds on previous artifacts

Next Module Preview

  • One clear sentence about next capability
  • Quote from a persona that teases the upcoming challenge
  • Shows progression of complexity or integration

Usage Guide

1. Research Phase (If URL Provided)

bash
# User provides URL for research
"@module-author Create module 7 on MCP servers using https://github.blog/mcp-announcement"

Research workflow:

  1. Read the provided URL content
  2. Extract key capabilities and features
  3. Identify how each persona would use this
  4. Note official documentation links
  5. Plan exercises that demonstrate value

2. Generate README.md

Using TEMPLATE.md, create:

  • Story So Far — Connection to previous modules
  • Why This Matters — Individual statements for ALL 5 personas
  • What You'll Learn — Brief capability overview
  • 📋 Exercise Planning Table — Detailed specifications for 2-3 exercises
  • What This Feature Does — Technical overview with official docs
  • Next Up — Preview of next module

3. Validate Against Checklist

  • README.md is ~100-150 lines
  • Story So Far connects to previous module's artifacts
  • Why This Matters includes ALL personas (Sarah, David, Marcus, Elena, Rafael)
  • Exercise planning table includes all required columns
  • Each exercise has:
    • Clear lead persona (⭐) and support personas (🤝)
    • Specific problem with current metrics
    • Brief solution description
    • Target improvement metrics (before→after format)
    • Artifacts that will be created
  • "What This Feature Does" includes 2-3 official documentation links
  • Next Up preview included
  • Section separators (---) between major sections

Example Invocation

code
@module-author Create Module 4 on custom instructions. Research https://docs.github.com/copilot/customizing-github-copilot/adding-custom-instructions-for-github-copilot

Plan these exercises:
- Sarah enforcing React component standards (saves 40 min/review)
- Marcus automating build validation (saves 28 min/failure)
- Elena generating comprehensive test suites (saves 43 min/suite)

Result:

  • modules/04-custom-instructions/README.md with:
    • Story connecting to Module 3's prompt libraries
    • Why This Matters for all 5 personas
    • Exercise planning table with 3 rows (detailed specifications)
    • Technical overview with official docs links
    • Next Up preview to Module 5

Next steps:

  • User invokes @exercise-author to generate exercise-01.md, exercise-02.md, exercise-03.md

Key Benefits of This Approach

Research Integration:

  • ✅ Learn new capabilities from official sources
  • ✅ Identify persona-relevant use cases
  • ✅ Find authoritative documentation to reference

Strategic Planning:

  • ✅ Module-level content strategy separate from exercise implementation
  • ✅ Clear specifications for exercise-author to implement
  • ✅ Ensures all personas are addressed at planning stage

Iterative Refinement:

  • ✅ Update module overview without touching exercises
  • ✅ Refine individual exercises independently
  • ✅ Add new exercises without regenerating README

Clear Separation:

  • ✅ module-author: Strategy, research, planning (README.md)
  • ✅ exercise-author: Tactical implementation (exercise-*.md files)