github · github-actions · Jun 23, 2026
@@ -3,7 +3,7 @@ title: 'Agents and Subagents'
 description: 'Learn how delegated subagents differ from primary agents, when to use them, and how to launch them in VS Code and Copilot CLI.'
 authors:
   - GitHub Copilot Learning Hub Team
-lastUpdated: 2026-05-07
+lastUpdated: 2026-06-23
 estimatedReadingTime: '9 minutes'
 tags:
   - agents
@@ -134,20 +134,52 @@ The important behavior is different from a single chat turn:
 
 That makes `/fleet` a practical way to launch subagents even if you are not authoring custom agent files yourself.
 
-### Rubber-duck agent (experimental)
+### Rubber-duck agent
 
-Available in `/experimental` (v1.0.42+), the **rubber-duck agent** applies a novel multi-model pattern: when you're working in a GPT-powered session, the rubber-duck agent internally routes certain requests through Claude to provide a second perspective. The idea is similar to rubber-duck debugging — talking through a problem with a different "listener" often surfaces assumptions or blind spots you didn't notice.
+The **rubber-duck agent** applies a novel multi-model pattern: when you're working in a GPT-powered session, the rubber-duck agent internally routes certain requests through Claude to provide a second perspective. The idea is similar to rubber-duck debugging — talking through a problem with a different "listener" often surfaces assumptions or blind spots you didn't notice.
 
-To try it, enable experimental features and then select the rubber-duck agent from the agent picker:
+As of v1.0.58, the rubber-duck agent is **enabled by default**. You can disable it with:
+
+```bash
+copilot config set builtInAgents.rubberDuck false
+```
+
+Or control whether it activates automatically without being explicitly invoked:
+
+```bash
+copilot config set builtInAgents.rubberDuckAutoInvoke false
+```
+
+To manually select it from the agent picker:
 
 ```
-/experimental           # toggle experimental features
 /agent                  # open the agent picker and select rubber-duck
 ```
 
 Because it runs as a sub-agent layer rather than replacing your primary model, you keep your current session model and context while the rubber-duck analysis runs in the background.
 
-> **Note**: This is an experimental feature and may change. Provide feedback via `/feedback` if you find it useful.
+### Configuring subagent model and reasoning (v1.0.62+)
+
+You can configure the default model, reasoning effort, and context tier for all built-in subagents via user settings or the `/subagents` picker (also accessible as `/agents`):
+
+- **Model**: Set a different AI model for subagent work (e.g., a faster model for research tasks)
+- **Reasoning effort**: Control how much reasoning budget subagents use (`low`, `medium`, `high`)
+- **Context tier**: Select the context window size subagents operate with
+
+Open the picker from within a session:
+
+```
+/subagents     # or /agents — opens the subagent configuration picker
+```
+
+Or configure via settings:
+
+```bash
+copilot config set subagents.model claude-sonnet-4
+copilot config set subagents.reasoningEffort low
+```
+
+This is particularly useful when you want fast, lightweight subagents for research and exploration, while keeping a higher-capability model for the main session.
 
 ## Orchestration patterns that work well
 

@@ -3,11 +3,12 @@ title: 'Automating with Hooks'
 description: 'Learn how to use hooks to automate lifecycle events like formatting, linting, and governance checks during Copilot agent sessions.'
 authors:
   - GitHub Copilot Learning Hub Team
-lastUpdated: 2026-05-08
+lastUpdated: 2026-06-23
 estimatedReadingTime: '8 minutes'
 tags:
   - hooks
   - automation
+  - github-actions
   - fundamentals
 relatedArticles:
   - ./building-custom-agents.md
@@ -182,7 +183,10 @@ Hooks support two types: `"command"` for running local shell scripts, and `"http
 
 **matcher** *(optional)*: A regular expression matched against the tool name. When present, the hook only fires for tools whose name fully matches the regex. For example, `"^bash$"` ensures the hook only runs for the `bash` tool, not for `edit` or other tools. This is particularly useful for `preToolUse` and `postToolUse` hooks where you want to target a specific tool.
 
-> **Important (v1.0.36+)**: Prior to v1.0.36, the `matcher` field was silently ignored — hooks with a `matcher` fired for all tool calls regardless of the regex. After upgrading to v1.0.36 or later, only tool calls whose name fully matches the `matcher` regex will trigger the hook. Review any existing `preToolUse`/`postToolUse` hooks that use `matcher` to ensure they still fire as expected.
+> **Important**: The `matcher` field behavior has been refined across versions:
+> - **Prior to v1.0.36**: `matcher` was silently ignored — hooks fired for all tool calls regardless of the regex.
+> - **v1.0.36+**: `preToolUse` matchers are correctly enforced.
+> - **v1.0.63+**: `postToolUse` matchers (e.g., `Edit|Write`) are now also correctly enforced. Before this version, `postToolUse` matchers were silently dropped, causing formatters and linters to fire after every tool call rather than only after edits. If you relied on `postToolUse` matchers to scope formatting hooks, upgrade to v1.0.63 or later.
 
 **cwd**: Working directory for the command (relative to repository root).
 

@@ -3,7 +3,7 @@ title: 'Installing and Using Plugins'
 description: 'Learn how to find, install, and manage plugins that extend GitHub Copilot CLI with reusable agents, skills, hooks, and integrations.'
 authors:
   - GitHub Copilot Learning Hub Team
-lastUpdated: 2026-04-27
+lastUpdated: 2026-06-23
 estimatedReadingTime: '8 minutes'
 tags:
   - plugins
@@ -33,9 +33,16 @@ A plugin bundles one or more of the following components:
 | **Hooks** | Event handlers that intercept agent behavior | `hooks.json` or `hooks/` |
 | **MCP Servers** | Model Context Protocol integrations for external tools | `.mcp.json` or `.github/mcp.json` |
 | **LSP Servers** | Language Server Protocol integrations | `lsp.json` or `.github/lsp.json` |
+| **Extensions** | Canvas or other session extensions bundled with the plugin | `extensions/` (v1.0.62+) |
 
 A plugin might include all of these or just one — for example, a plugin could provide a single specialized agent, or an entire development toolkit with multiple agents, skills, hooks, and MCP server configurations working together.
 
+### Extensions in Plugins (v1.0.62+)
+
+Starting with v1.0.62, plugins can ship **extensions** — canvas providers or other session extensions that become available immediately when the plugin is installed. This means a plugin can not only provide agents and skills but also add new canvas types or interactive surfaces to your Copilot sessions.
+
+Extensions bundled in plugins are loaded automatically alongside other plugin components — no extra configuration needed beyond installing the plugin.
+
 ### Example: What a Plugin Looks Like
 
 Here's the structure of a typical plugin:

@@ -3,7 +3,7 @@ title: 'Understanding MCP Servers'
 description: 'Learn how Model Context Protocol servers extend GitHub Copilot with access to external tools, databases, and APIs.'
 authors:
   - GitHub Copilot Learning Hub Team
-lastUpdated: 2026-05-07
+lastUpdated: 2026-06-23
 estimatedReadingTime: '8 minutes'
 tags:
   - mcp
@@ -66,6 +66,7 @@ MCP servers are configured per-workspace. GitHub Copilot CLI discovers server de
 | File | Scope | Notes |
 |------|-------|-------|
 | `.mcp.json` | Repository root | Preferred for repo-shared configuration |
+| `.github/mcp.json` | Repository root (alternate) | Auto-loaded from the `.github/` directory (v1.0.61+) |
 | `.vscode/mcp.json` | VS Code workspace | VS Code–compatible workspace config |
 | `devcontainer.json` | Dev container | Available when running inside a container |
 
@@ -119,6 +120,18 @@ This guided flow is the recommended way to add new MCP servers, especially for s
 
 **type** (remote servers): The transport type for remote MCP servers (`http` or `sse`). This field can now be omitted — the CLI defaults to `http` when no type is specified, simplifying remote server configuration.
 
+**deferTools** (v1.0.63+): When set to `true`, keeps the server's tools always available in the agent's context even when tool search is enabled. By default, tool search can hide infrequently-used tools to reduce context overhead; `deferTools` overrides this and ensures the server's tools are never hidden.
+
+```json
+{
+  "my-always-on-server": {
+    "command": "npx",
+    "args": ["-y", "@my-org/mcp-server"],
+    "deferTools": true
+  }
+}
+```
+
 ### Managing Persistent MCP Configuration via Server RPCs
 
 In addition to file-based configuration, GitHub Copilot CLI exposes **server RPCs** that let MCP servers and tooling scripts manage the persistent MCP server registry at runtime. This enables programmatic setup — for example, an installer script that registers a server without requiring you to hand-edit a JSON file.

@@ -3,7 +3,7 @@ title: 'Using the Copilot Coding Agent'
 description: 'Learn how to use GitHub Copilot coding agent to autonomously work on issues, generate pull requests, and automate development tasks.'
 authors:
   - GitHub Copilot Learning Hub Team
-lastUpdated: 2026-05-13
+lastUpdated: 2026-06-23
 estimatedReadingTime: '12 minutes'
 tags:
   - coding-agent
@@ -138,12 +138,18 @@ Custom agents let you give the coding agent a specialized persona, toolset, and
 
 **Where custom agents live**: Agent profiles are stored as `.agent.md` files in `.github/agents/` in your repository. For organization-wide agents, place them in the root `agents/` directory.
 
+> **Tip (v1.0.62+)**: Custom agents in **nested** `.github/agents/` or `.claude/agents/` directories are automatically discovered when a session is started from a subdirectory of the repository root. This is useful for monorepos where each package or service can have its own tailored agents without polluting the repository-root agents directory.
+
 ```
 .github/
 └── agents/
     ├── api-architect.agent.md
     ├── test-specialist.agent.md
     └── security-reviewer.agent.md
+
+# Also discovered when starting from packages/backend/:
+packages/backend/.github/agents/
+    └── backend-specialist.agent.md
 ```
 
 **Selecting an agent on GitHub.com**: When prompting the coding agent or assigning it to an issue, use the dropdown menu in the agents panel to select your custom agent instead of the default.