Document pool rate limiting in Copilot instructions

.github/copilot-instructions.md

@@ -150,6 +150,24 @@ When a new issue is created, follow these steps:
 - Do not modify the `CODEOWNERS` file directly.
 - Do not modify `CHANGELOG.md` unless executing a release workflow (see `release-notes` prompt).
 - Do not close issues without a fix or without providing a clear reason.
+- **Check `.github/instructions/` for matching `applyTo` patterns BEFORE starting any file creation or modification task.** Apply all conventions from those files from the start—do not refactor later.
+
+## Instruction File Lookup (Required Workflow)
+
+Before creating or modifying any file, check if matching instruction files exist:
+
+1. Scan `.github/instructions/` for files with `applyTo` patterns that match your file path
+2. Read **all** matching instruction files completely before writing code
+3. Apply **all** conventions, patterns, and requirements from the start
+4. Do not skip instructions or plan to refactor later
+
+**Example:** Creating `src/Microsoft.Data.SqlClient/tests/UnitTests/ConnectionPool/NewTest.cs`
+- File matches `applyTo: "**/tests/**,**/*Test*.cs"` from `testing.instructions.md`
+- Must read `testing.instructions.md` before writing the first test method
+- Must apply AAA pattern, XML doc comments, and all testing conventions from the start
+- Do not create tests first and add documentation later
+
+This ensures consistency across the codebase and prevents rework.
 
 ## Terminal Execution Safety
 - Treat any non-zero shell exit code as a failed step that requires correction before proceeding.

.github/instructions/testing.instructions.md

@@ -189,10 +189,15 @@ class and method level.
 #### What the Comments Must Explain
 - The behavior/contract being tested (not just restating the method name).
 - Why the scenario matters (for example: regression guard, parsing contract, sync/async parity,
-  isolation requirement).
+  isolation requirement, thread-safety validation, edge case handling).
 - For helper methods, what side effects occur (for example console redirection, file system
   copying, process execution) and why they are needed.
 
+#### Implementation Guidance
+- Always use the **Arrange-Act-Assert (AAA)** pattern with explicit section comments (see [Test Structure](#test-structure-arrange-act-assert-aaa-pattern)).
+- Keep the test body focused on a single logical assertion; avoid testing multiple unrelated behaviors.
+- Extract complex setup into helper methods or fixtures rather than embedding it in the test.
+
 #### Style Guidance
 - Keep comments concise and factual.
 - Prefer behavior-focused wording over implementation trivia.
@@ -212,25 +217,41 @@ public void AppRunWithMalformedConnectionStringReturnsOneAndWritesParseError()
 }
 ```
 
-### Test Structure
+### Test Structure: Arrange-Act-Assert (AAA) Pattern
+
+All test methods **must** follow the **Arrange-Act-Assert (AAA)** pattern with explicit section comments.
+This pattern improves readability and maintainability by clearly delineating setup, execution, and validation.
+
+**Pattern:**
 ```csharp
-public class FeatureNameTests
+[Fact]
+public void MethodName_Scenario_ExpectedResult()
 {
-    [Fact]
-    public void MethodName_Scenario_ExpectedResult()
-    {
-        // Arrange
-        var sut = new SystemUnderTest();
-
-        // Act
-        var result = sut.PerformAction();
-
-        // Assert
-        Assert.Equal(expected, result);
-    }
+    // Arrange
+    // Set up test fixtures, initial state, dependencies, and test data
+    var sut = new SystemUnderTest();
+    var input = new TestData();
+
+    // Act
+    // Execute the code under test
+    var result = sut.PerformAction(input);
+
+    // Assert
+    // Validate outcomes and expectations
+    Assert.Equal(expected, result);
 }
 ```
 
+**Benefits:**
+- Clear visual separation of setup, execution, and validation
+- Easier to identify test logic flow
+- Simpler to debug failing tests (know which section failed)
+- Consistent convention across codebase
+- Improves AI agent comprehension and test generation
+
+**Guideline:** Each section should be visually distinct. If AAA sections become too large (e.g., Arrange > 20 lines),
+consider extracting helper methods or fixtures rather than embedding complexity in the test method itself.
+
 ### Naming Conventions
 - Test class: `{ClassName}Tests`
 - Test method: `{Method}_{Scenario}_{ExpectedBehavior}`