docs: add .claude/settings.json support and model picker reasoning effort

- Chapter 04: Add .claude/settings.json and .claude/settings.local.json
  to the Instruction File Formats table, plus a new 'Settings Files' section
  explaining how they differ from instruction files (added in CLI v1.0.12)
- Chapter 01: Update the Switching Models section to describe the new
  full-screen model picker and inline reasoning effort adjustment with
  ← / → arrow keys (added in CLI v1.0.12)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: DanWahlin

01-setup-and-first-steps/README.md

@@ -464,11 +464,13 @@ Copilot CLI supports multiple AI models from OpenAI, Anthropic, Google, and othe
 copilot
 > /model
 
-# Shows available models and lets you pick one. Select Sonnet 4.5.
+# Opens a full-screen model picker. Select a model and press Enter.
 ```
 
 > 💡 **Tip**: Some models cost more "premium requests" than others. Models marked **1x** (like Claude Sonnet 4.5) are a great default. They're capable and efficient. Higher-multiplier models use your premium request quota faster, so save those for when you really need them.
 
+> 💡 **Reasoning effort**: Inside the model picker, use the ← and → arrow keys to adjust the **reasoning effort** for the selected model. Lower effort is faster; higher effort is more thorough. The current effort level is shown next to the model name in the session header (e.g., `claude-sonnet-4.5 (high)`). Start with the default and raise it only when you need deeper analysis.
+
 </details>
 
 ---

04-agents-custom-instructions/README.md

@@ -423,9 +423,27 @@ Copilot will scan your project and create tailored instruction files. You can ed
 | `.github/copilot-instructions.md` | Project | GitHub Copilot specific |
 | `.github/instructions/*.instructions.md` | Project | Granular, topic-specific instructions |
 | `CLAUDE.md`, `GEMINI.md` | Project root | Supported for compatibility |
+| `.claude/settings.json` | Project | Project-level behavior settings (committed, shared with team) |
+| `.claude/settings.local.json` | Project | Personal settings override (not committed, stays on your machine) |
 
 > 🎯 **Just getting started?** Use `AGENTS.md` for project instructions. You can explore the other formats later as needed.
 
+### Settings Files (.claude/settings.json)
+
+In addition to instruction files, Copilot reads **settings files** from your project's `.claude/` folder. These control *how* Copilot behaves rather than *what it knows* about your project. Think of instruction files as "here's our coding standards" and settings files as "here's how the tool should behave."
+
+- **`.claude/settings.json`** — Project settings committed to your repo, shared with the whole team.
+- **`.claude/settings.local.json`** — Your personal overrides, not committed (add it to `.gitignore`).
+
+```
+your-project/
+└── .claude/
+    ├── settings.json        # committed — everyone gets these defaults
+    └── settings.local.json  # NOT committed — personal overrides
+```
+
+> 💡 **Tip**: Both files are optional. Start with `AGENTS.md` for instructions, and only add a `settings.json` when you need to tune Copilot's behavior for the whole team.
+
 ### AGENTS.md
 
 `AGENTS.md` is the recommended format. It's an [open standard](https://agents.md/) that works across Copilot and other AI coding tools. Place it in your repository root and Copilot reads it automatically. This project's own [AGENTS.md](../AGENTS.md) is a working example.