feat(testing): redesign testing architecture, resolve performance warnings, and update guide

[cesarcastrocuba]

This Pull Request introduces a comprehensive restructuring of the framework's testing system, moving toward a more scalable, decoupled architecture aligned with modern .NET best practices. Additionally, it addresses multiple critical performance warnings and synchronizes the development environment's testing guide.

Key Changes:

1. Testing Architecture Redesign:

   • Transitioned from a monolithic test model to a specialized modular structure:
     • Architecture.Tests: Enforces layering and dependency rules.
     • Shared.Tests: Common infrastructure and container configuration.
     • Integration.Tests: Persistence and business logic (MediatR) testing without HTTP.
     • Functional.Tests: End-to-End vertical slices via HTTP.
     • Spec.Tests: BDD Acceptance specs using Gherkin.
   • Elimination of InMemoryDatabase: Switched to Testcontainers (PostgreSQL/Redis) to ensure environment parity and eliminate "it works on my machine" issues.
   • Mocking Standardization: Transitioned from Moq to NSubstitute for cleaner syntax and improved maintainability.

2. Performance & Code Quality:

   • Logging Performance (CA1873): Wrapped expensive logging calls with _logger.IsEnabled(LogLevel.Information) guards across the Identity, Multitenancy, and Auditing modules. This prevents unnecessary allocations and string concatenations when the logging level is disabled.
   • Dependency Updates: Updated MimeKit/MailKit to resolve vulnerabilities and improved nullability handling in mail services.

3. Documentation & Skills Synchronization:

   • Fully updated the testing guide in .claude/skills/testing-guide/SKILL.md.
   • Restored structural references for module-specific test projects (Identity, Multitenancy, Auditing, Generic) to ensure the AI assistant remains aware of the full test suite.

4. Functional Verification Fixes:

   • Resolved Identity endpoint path mismatches.
   • Fixed TokenResponse typing issues that were causing JsonElement deserialization errors during functional test execution.

Verification Results:

• Build Status: 0 Errors, 0 Critical Performance Warnings.
• Test Execution: 465+ tests passing across all projects.
• Process: The framework is now fully consistent with the Spec-Driven Design workflow.