Add .github/agents documentation and tour steps across all tracks

Typescript/.tours/typscript-ghcp-practice.tour

@@ -139,6 +139,11 @@
       "title": "Create a Custom Chat Mode: PRD builder",
       "description": "Let's set up our own custom GitHub Copilot Chat Mode to accelerate creation of Product Requirements Documents (PRDs).\n\n**Goal**: Create a mode named **PRD builder** that loads the persona instructions so you can quickly draft structured PRDs from short prompts.\n\n### Steps\n1. Open the Copilot Chat panel (Ctrl + Alt + I).\n1. Click the mode dropdown (where you pick *Ask* / *Agent*). You probably noticed there are already two custom modes, let's create our own by selecting **Configure modes** (see screenshot below).\n\n    ![Configure Modes](./assets/ConfigureModes.png)\n\n1. In the configuration UI, click **Create a new custom mode**.\n\n    ![Create Mode](./assets/Create a new custom mode.png)\n\n1. Let's use the User level scope to share this across projects instead of being project specific\n\n    ![User Data Folder](./assets/UserDataFolder.png)    \n        \n1. Name it exactly: `PRD builder`.\n1. Open this file `PRD-Builder-Persona.md` ([link](./personas/PRD-Builder-Persona.md)) and copy its entire contents into the new mode's definition/instructions text area.\n1. (Optional) Pick an emoji/icon to make it stand out (🧱, 📄, 🚀).\n1. Save the mode.\n\nCongratulations, you've created a custom mode.\n\n\n### What This Gives You\n- Consistent PRD structure (intro, goals, user stories, requirements, non-goals, metrics)\n- Automatic reminder to ask clarifying questions first\n- Faster onboarding for teammates who can reuse the same mode\n\n### Pro Tips\n- You can store multiple personas (e.g., *Test Plan Builder*, *API Spec Writer*) the same way.\n- Version your persona files in Git so the team can improve them collaboratively.\n- If you update the persona text here, re-open the mode config and re-paste to sync.\n\nWhen you're done, try giving it a *very short* feature idea and see how it first asks you clarifying questions (that's by design!).\n\nGive it a spin—your future specs will thank you. ✨"
     },
+    {
+      "file": "Agents.md",
+      "title": "Understanding .github/agents: Repository-Level AI Assistants 🤖",
+      "description": "**Discover the Power of GitHub Copilot Agents!**\n\nNow that you've created custom chat modes, let's explore an even more powerful feature: **repository-level AI agents** stored in `.github/agents/`.\n\n### What Makes .github/agents Special?\n\n1. **Multiple AI Assistants Out of the Box** 🎯\n   - Visit [agents.md](https://agents.md) to see all the AI assistants GitHub Copilot supports natively!\n   - Each agent is specialized for different tasks: code review, documentation, testing, architecture, and more\n   - No extra configuration needed - just reference them in your Copilot chat\n\n2. **Nested Folder Support** 📁\n   - Organize agents in subdirectories: `.github/agents/backend/`, `.github/agents/frontend/`, etc.\n   - **Deeper folders = richer context**: When working on backend code, backend-specific agents automatically have better context\n   - Example structure:\n     ```\n     .github/agents/\n     ├── tour-agent.md          # General tour building\n     ├── backend/\n     │   └── api-reviewer.md    # Backend API reviews\n     └── frontend/\n         └── ui-helper.md       # Frontend component help\n     ```\n\n3. **Team Collaboration** 👥\n   - Agents are version-controlled alongside your code\n   - Different teams can maintain their own specialized agents\n   - Easy to share knowledge and best practices across the organization\n\n### Try It Now!\n\n1. **Explore this repo's agent**: Open `.github/agents/tour-agent.md` to see the agent that helped build this tour\n2. **Visit agents.md**: Check out [https://agents.md](https://agents.md) to discover all available AI assistants\n3. **Think about your needs**: What agents would help your team? Code reviewers? Test generators? Documentation writers?\n\n### Key Difference: Chat Modes vs Agents\n\n- **Chat Modes** (`.github/chatmodes/`): User-specific customizations, great for personal workflows\n- **Agents** (`.github/agents/`): Repository-level assistants, perfect for team-wide standards and practices\n\nBoth work together to give you maximum flexibility! 🚀\n\n**Pro Tip**: Combine agents with Copilot Instructions (`.github/copilot-instructions.md`) for the ultimate context-aware AI assistance!"
+    },
     {
       "file": "EasterEggs.md",
       "description": "From this list of emails below, I only want the non-domain part formatted with new line separators",
@@ -151,4 +156,4 @@
     }
   ],
   "ref": "main"
-}
+}

Typescript/Agents.md

@@ -0,0 +1,58 @@
+# Understanding GitHub Copilot Agents
+
+## What are GitHub Copilot Agents?
+
+GitHub Copilot Agents are specialized AI assistants that can be configured to help with specific development tasks. They live in your repository's `.github/agents/` directory and can be customized to understand your project's context, coding standards, and workflows.
+
+## Key Features
+
+### 🤖 Multiple AI Assistants Out of the Box
+When you navigate to [agents.md](https://agents.md), you'll discover a variety of pre-built AI agents that GitHub Copilot supports right away! These agents can:
+- Help with code reviews
+- Generate documentation
+- Assist with testing strategies
+- Provide architectural guidance
+- And much more!
+
+### 📁 Nested Folder Support
+Agents can be organized in nested folders within `.github/agents/` to provide deeper, more context-specific assistance:
+
+```
+.github/
+└── agents/
+    ├── tour-agent.md              # Top-level tour building agent
+    ├── backend/
+    │   ├── api-reviewer.md        # Backend-specific API review agent
+    │   └── database-helper.md     # Database query assistance
+    └── frontend/
+        ├── ui-component-agent.md  # UI component generation
+        └── accessibility-agent.md # Accessibility checking
+```
+
+This hierarchical structure allows you to:
+- **Organize by domain** - Separate concerns (backend, frontend, infrastructure)
+- **Enrich context** - Deeper folders provide more specific context when working on specialized code
+- **Team collaboration** - Different teams can maintain their own agent configurations
+
+## How to Use Agents
+
+1. **Discover available agents** - Visit [agents.md](https://agents.md) to explore supported AI assistants
+2. **Create custom agents** - Add `.md` files in `.github/agents/` with your agent instructions
+3. **Reference in chat** - Use agents through GitHub Copilot chat with specific commands
+4. **Iterate and improve** - Update agent instructions based on team feedback and evolving needs
+
+## Example: Tour Agent
+
+This repository includes a tour-building agent at `.github/agents/tour-agent.md` that helps create interactive CodeTour experiences for learning GitHub Copilot features.
+
+## Best Practices
+
+- **Keep agents focused** - Each agent should have a specific, well-defined purpose
+- **Use nested folders** - Organize agents by domain or feature area for better context
+- **Document agent capabilities** - Clearly describe what each agent can help with
+- **Version control** - Track agent changes like any other code asset
+- **Team review** - Have team members review and improve agent instructions together
+
+---
+
+**Pro Tip**: Custom agents work best when combined with GitHub Copilot Instructions (`.github/copilot-instructions.md`) to provide comprehensive context about your project's coding standards and preferences!

cpp/.tours/advanced-ghcp-lab-for-cpp.tour

@@ -186,6 +186,11 @@
       "title": "Create a Custom Chat Mode: PRD builder",
       "description": "Let's set up our own custom GitHub Copilot Chat Mode to accelerate creation of Product Requirements Documents (PRDs).\n\n**Goal**: Create a mode named **PRD builder** that loads the persona instructions so you can quickly draft structured PRDs from short prompts.\n\n### Steps\n1. Open the Copilot Chat panel (Ctrl + Alt + I).\n1. Click the mode dropdown (where you pick *Ask* / *Agent*). You probably noticed there are already two custom modes, let's create our own by selecting **Configure modes** (see screenshot below).\n\n    ![Configure Modes](./assets/ConfigureModes.png)\n\n1. In the configuration UI, click **Create a new custom mode**.\n\n    ![Create Mode](./assets/Create a new custom mode.png)\n\n1. Let's use the User level scope to share this across projects instead of being project specific\n\n    ![User Data Folder](./assets/UserDataFolder.png)    \n        \n1. Name it exactly: `PRD builder`.\n1. Open this file `PRD-Builder-Persona.md` ([link](./personas/PRD-Builder-Persona.md)) and copy its entire contents into the new mode's definition/instructions text area.\n1. (Optional) Pick an emoji/icon to make it stand out (🧱, 📄, 🚀).\n1. Save the mode.\n\nCongratulations, you've created a custom mode.\n\n\n### What This Gives You\n- Consistent PRD structure (intro, goals, user stories, requirements, non-goals, metrics)\n- Automatic reminder to ask clarifying questions first\n- Faster onboarding for teammates who can reuse the same mode\n\n### Pro Tips\n- You can store multiple personas (e.g., *Test Plan Builder*, *API Spec Writer*) the same way.\n- Version your persona files in Git so the team can improve them collaboratively.\n- If you update the persona text here, re-open the mode config and re-paste to sync.\n\nWhen you're done, try giving it a *very short* feature idea and see how it first asks you clarifying questions (that's by design!).\n\nGive it a spin—your future specs will thank you. ✨"
     },
+    {
+      "file": "src/Agents.md",
+      "title": "Understanding .github/agents: Repository-Level AI Assistants 🤖",
+      "description": "**Discover the Power of GitHub Copilot Agents!**\n\nNow that you've created custom chat modes, let's explore an even more powerful feature: **repository-level AI agents** stored in `.github/agents/`.\n\n### What Makes .github/agents Special?\n\n1. **Multiple AI Assistants Out of the Box** 🎯\n   - Visit [agents.md](https://agents.md) to see all the AI assistants GitHub Copilot supports natively!\n   - Each agent is specialized for different tasks: code review, documentation, testing, architecture, and more\n   - No extra configuration needed - just reference them in your Copilot chat\n\n2. **Nested Folder Support** 📁\n   - Organize agents in subdirectories: `.github/agents/backend/`, `.github/agents/frontend/`, etc.\n   - **Deeper folders = richer context**: When working on backend code, backend-specific agents automatically have better context\n   - Example structure:\n     ```\n     .github/agents/\n     ├── tour-agent.md          # General tour building\n     ├── backend/\n     │   └── api-reviewer.md    # Backend API reviews\n     └── frontend/\n         └── ui-helper.md       # Frontend component help\n     ```\n\n3. **Team Collaboration** 👥\n   - Agents are version-controlled alongside your code\n   - Different teams can maintain their own specialized agents\n   - Easy to share knowledge and best practices across the organization\n\n### Try It Now!\n\n1. **Explore this repo's agent**: Open `.github/agents/tour-agent.md` to see the agent that helped build this tour\n2. **Visit agents.md**: Check out [https://agents.md](https://agents.md) to discover all available AI assistants\n3. **Think about your needs**: What agents would help your team? Code reviewers? Test generators? Documentation writers?\n\n### Key Difference: Chat Modes vs Agents\n\n- **Chat Modes** (`.github/chatmodes/`): User-specific customizations, great for personal workflows\n- **Agents** (`.github/agents/`): Repository-level assistants, perfect for team-wide standards and practices\n\nBoth work together to give you maximum flexibility! 🚀\n\n**Pro Tip**: Combine agents with Copilot Instructions (`.github/copilot-instructions.md`) for the ultimate context-aware AI assistance!"
+    },
     {
       "file": "src/Program.cpp",
       "description": "Congratulations!!!\r\n\r\n\r\nYou have completed the lab, I hope that you have learned a thing or two and had some fun while you're at it.\r\n\r\nRemember - \r\n\r\n\r\nShort.\r\n\r\n\r\nSimple.\r\n\r\n\r\nSpecific.\r\n\r\n\r\n\r\n\r\nAnd provide the relevant context (#file/ #selection, etc.)",

cpp/src/Agents.md

@@ -0,0 +1,58 @@
+# Understanding GitHub Copilot Agents
+
+## What are GitHub Copilot Agents?
+
+GitHub Copilot Agents are specialized AI assistants that can be configured to help with specific development tasks. They live in your repository's `.github/agents/` directory and can be customized to understand your project's context, coding standards, and workflows.
+
+## Key Features
+
+### 🤖 Multiple AI Assistants Out of the Box
+When you navigate to [agents.md](https://agents.md), you'll discover a variety of pre-built AI agents that GitHub Copilot supports right away! These agents can:
+- Help with code reviews
+- Generate documentation
+- Assist with testing strategies
+- Provide architectural guidance
+- And much more!
+
+### 📁 Nested Folder Support
+Agents can be organized in nested folders within `.github/agents/` to provide deeper, more context-specific assistance:
+
+```
+.github/
+└── agents/
+    ├── tour-agent.md              # Top-level tour building agent
+    ├── backend/
+    │   ├── api-reviewer.md        # Backend-specific API review agent
+    │   └── database-helper.md     # Database query assistance
+    └── frontend/
+        ├── ui-component-agent.md  # UI component generation
+        └── accessibility-agent.md # Accessibility checking
+```
+
+This hierarchical structure allows you to:
+- **Organize by domain** - Separate concerns (backend, frontend, infrastructure)
+- **Enrich context** - Deeper folders provide more specific context when working on specialized code
+- **Team collaboration** - Different teams can maintain their own agent configurations
+
+## How to Use Agents
+
+1. **Discover available agents** - Visit [agents.md](https://agents.md) to explore supported AI assistants
+2. **Create custom agents** - Add `.md` files in `.github/agents/` with your agent instructions
+3. **Reference in chat** - Use agents through GitHub Copilot chat with specific commands
+4. **Iterate and improve** - Update agent instructions based on team feedback and evolving needs
+
+## Example: Tour Agent
+
+This repository includes a tour-building agent at `.github/agents/tour-agent.md` that helps create interactive CodeTour experiences for learning GitHub Copilot features.
+
+## Best Practices
+
+- **Keep agents focused** - Each agent should have a specific, well-defined purpose
+- **Use nested folders** - Organize agents by domain or feature area for better context
+- **Document agent capabilities** - Clearly describe what each agent can help with
+- **Version control** - Track agent changes like any other code asset
+- **Team review** - Have team members review and improve agent instructions together
+
+---
+
+**Pro Tip**: Custom agents work best when combined with GitHub Copilot Instructions (`.github/copilot-instructions.md`) to provide comprehensive context about your project's coding standards and preferences!

dotNet/.tours/advanced-ghcp-lab-for-c.tour

@@ -287,6 +287,11 @@
         }
       }
     },
+    {
+      "file": "src/Agents.md",
+      "title": "Understanding .github/agents: Repository-Level AI Assistants 🤖",
+      "description": "**Discover the Power of GitHub Copilot Agents!**\n\nNow that you've created custom chat modes, let's explore an even more powerful feature: **repository-level AI agents** stored in `.github/agents/`.\n\n### What Makes .github/agents Special?\n\n1. **Multiple AI Assistants Out of the Box** 🎯\n   - Visit [agents.md](https://agents.md) to see all the AI assistants GitHub Copilot supports natively!\n   - Each agent is specialized for different tasks: code review, documentation, testing, architecture, and more\n   - No extra configuration needed - just reference them in your Copilot chat\n\n2. **Nested Folder Support** 📁\n   - Organize agents in subdirectories: `.github/agents/backend/`, `.github/agents/frontend/`, etc.\n   - **Deeper folders = richer context**: When working on backend code, backend-specific agents automatically have better context\n   - Example structure:\n     ```\n     .github/agents/\n     ├── tour-agent.md          # General tour building\n     ├── backend/\n     │   └── api-reviewer.md    # Backend API reviews\n     └── frontend/\n         └── ui-helper.md       # Frontend component help\n     ```\n\n3. **Team Collaboration** 👥\n   - Agents are version-controlled alongside your code\n   - Different teams can maintain their own specialized agents\n   - Easy to share knowledge and best practices across the organization\n\n### Try It Now!\n\n1. **Explore this repo's agent**: Open `.github/agents/tour-agent.md` to see the agent that helped build this tour\n2. **Visit agents.md**: Check out [https://agents.md](https://agents.md) to discover all available AI assistants\n3. **Think about your needs**: What agents would help your team? Code reviewers? Test generators? Documentation writers?\n\n### Key Difference: Chat Modes vs Agents\n\n- **Chat Modes** (`.github/chatmodes/`): User-specific customizations, great for personal workflows\n- **Agents** (`.github/agents/`): Repository-level assistants, perfect for team-wide standards and practices\n\nBoth work together to give you maximum flexibility! 🚀\n\n**Pro Tip**: Combine agents with Copilot Instructions (`.github/copilot-instructions.md`) for the ultimate context-aware AI assistance!"
+    },
     {
       "title": "GitHub Copilot Coding Agent: Official Demo & Repo",
       "description": "Let's take it one step further and explore the official GitHub Copilot Agent (ADO) project!\n\nThe Copilot Agent is an open-source, extensible framework for building your own AI-powered developer workflows—beyond just chat and completions.\n\n**What can you do with it?**\n- Integrate with GitHub\n- Automate code reviews, PR triage, and more\n- Build custom agents for your team's unique needs\n\n**Want to understand the difference?**\n📖 [Learn about the differences between coding agent and agent mode](https://github.blog/developer-skills/github/less-todo-more-done-the-difference-between-coding-agent-and-agent-mode-in-github-copilot/) - This blog post explains when to use each approach!\n\n**Demo time:**\n1. Let's visit the [GHCP-Lab](https://github.com/microsoft/GHCP-Lab) and try it out.\n\nThis is where the future of AI-powered development is headed—give it a try and imagine the possibilities! 🚀"
@@ -297,4 +302,4 @@
       "title": "Your Copilot Journey Starts Now! 🚀"
     }
   ]
-}
+}