Fix 5 inaccuracies found in tutorial content review

Co-authored-by: matthewwilson <7092887+matthewwilson@users.noreply.github.com>

Author: Copilot

exercises/sample-app/package.json

@@ -1,7 +1,7 @@
 {
   "name": "task-manager",
   "version": "1.0.0",
-  "description": "A simple task management utility — used as a practice codebase for the Claude Code tutorial",
+  "description": "A simple task management utility — used as a practice codebase for the GitHub Copilot CLI tutorial",
   "private": true,
   "type": "module",
   "devDependencies": {

tutorial/lessons/08-writing-and-running-tests.md

@@ -41,7 +41,7 @@ Things to think about:
 1. Look at the existing tests in `tasks.test.ts` — notice the `describe`/`it` structure and how `beforeEach` resets state
 2. Include verification in your prompt — ask Copilot to both write the tests and run them: "Add tests for `deleteTask` in the sample app's test file, then run `npm test` to make sure they pass"
 3. Watch the test output — Copilot should show you the results of `npm test` so you can see the tests running
-4. If any tests fail, that is a teaching moment — it might mean the lesson 5 bug has not been fixed yet
+4. If any tests fail, that is a teaching moment — it might mean the lesson 6 bug has not been fixed yet
 
 ## Completion Criteria
 

tutorial/lessons/18-graduation.md

@@ -12,7 +12,7 @@ You have worked through the core of Copilot CLI hands-on:
 - Wrote and ran tests using Copilot's shell access
 - Created a custom agent for your own workflow
 - Watched Copilot delegate work to specialised agents
-- Learned how hooks automate workflows behind the scenes
+- Built reusable agent skills to give Copilot on-demand capabilities
 
 ## Key Takeaways
 
@@ -54,7 +54,7 @@ If you hit these problems in real work, here's where to look:
 
 | Frustration | Fix | Lesson |
 |---|---|---|
-| "Copilot forgets things between sessions" | Put standing instructions in copilot-instructions.md. Use `--continue` or `/resume` to pick up where you left off. | L3-4, L12 |
+| "Copilot forgets things between sessions" | Put standing instructions in copilot-instructions.md. Use `--continue` or `/resume` to pick up where you left off. | L3-4, L13 |
 | "Copilot changes too much / goes rogue" | Scope your prompts explicitly. Switch to Normal mode (use Shift+Tab to cycle modes) to review each edit. Use `/diff` to check changes. | L5, L12 |
 | "The code is too complex or over-engineered" | Add style rules to copilot-instructions.md ("prefer simple solutions", "no unnecessary abstractions"). Tell Copilot "write the simplest implementation." | L3-4, L5 |
 | "Endless back-and-forth refinement" | Stop. `/clear`. Use the "interview me" pattern to write a spec first. Put it in instructions. Start fresh. | L2, L7 |

tutorial/lessons/23-qa-track-graduation.md

@@ -27,7 +27,7 @@ Ask Copilot to generate a comprehensive test suite covering ALL remaining untest
 
 ### 2. QA Auditor Agent
 
-Create a custom agent at `.github/agents/qa-audit/AGENT.md` that tells Copilot to perform a QA audit: read all source files, check test coverage, identify risks, and produce a QA report. This is a practical example of building tools for your QA workflow.
+Create a custom agent at `.github/agents/qa-audit.agent.md` that tells Copilot to perform a QA audit: read all source files, check test coverage, identify risks, and produce a QA report. This is a practical example of building tools for your QA workflow.
 
 ### 3. CI Pipeline Design