test: Add comprehensive VS Code behavior tests with evidence-based proof

Implements 14 new tests that execute VS Code's ACTUAL organize imports
command to document exactly what it can and cannot do. This provides
executable proof for our marketing claims instead of speculation.

Key findings and test coverage:
- VS Code does NOT automatically separate external vs internal imports
- Proven: Mixed @angular/* and ./services/* sorted together as ONE group
- Comments within imports treated as group separators (edge case)
- Side-effect imports (import 'zone.js') sorted to END
- All duplicate merging and unused removal behaviors documented

Technical implementation:
- Execute real editor.action.organizeImports command (not mocked)
- Create temp files in os.tmpdir() (VS Code can organize standalone files)
- Requires 2-second delay for TypeScript language service analysis
- Requires 1-second delay after command (async execution)
- Total: ~3 seconds per test (acceptable for 14 tests)
- Uses shared test helpers (createTempDocument, deleteTempDocument)

Documentation improvements:
- Rewrote README.md comparison section (removed duplicates)
- Rewrote blog-post.md comparison section (clearer messaging)
- Emphasized automatic external/internal grouping as primary value
- Clarified pattern-based grouping is optional enhancement
- Added concrete examples showing VS Code's flat alphabetical sort

Updated audit documentation:
- Added comprehensive VS Code behavior tests section
- Documented test strategy, findings, and edge cases
- Listed 6 questions for reviewer feedback
- Included code examples and test evidence

Cleanup:
- Removed redundant onCommand activation events from package.json
  (VS Code auto-generates these from contributes.commands)
- Kept only onLanguage activation events

Test results: 309 tests passing (14 new), 0 failing

This change provides verifiable proof that automatic external/internal
import grouping is our #1 differentiator over VS Code's built-in feature.

Author: JohannesHoppe

.claude/settings.local.json

@@ -1,7 +1,23 @@
 {
   "permissions": {
     "allow": [
-      "Bash(node -e \"\nconst { Project } = require(''ts-morph'');\nconst project = new Project({ useInMemoryFileSystem: true });\nconst sf = project.createSourceFile(''test.ts'', \\`import data from ''./data.json'' with { type: ''json'' };\\`);\nconst imp = sf.getImportDeclarations()[0];\nconst allMethods = Object.getOwnPropertyNames(Object.getPrototypeOf(imp));\nconst attributeMethods = allMethods.filter(m => m.toLowerCase().includes(''assert'') || m.toLowerCase().includes(''attribute'') || m.toLowerCase().includes(''with''));\nconsole.log(''Import Attribute Related Methods:'', attributeMethods);\nconsole.log(''\\nAll methods count:'', allMethods.length);\nconsole.log(''Sample methods:'', allMethods.slice(0, 30));\n\")"
+      "Bash(npm run compile)",
+      "Bash(find src/test -name \"*.test.ts\" -exec sed -i '' 's/public insertSemicolons(_resource: Uri): boolean {/public async insertSemicolons(_resource: Uri): Promise<boolean> {/g' {} ;)",
+      "Bash(find src/test -name \"*.test.ts\" -exec sed -i '' 's/public stringQuoteStyle(_resource: Uri): .* {/public async stringQuoteStyle(_resource: Uri): Promise<\"\"\\\"\"\"\" | \"\"'\"''\"'\"\"> {/g' {} ;)",
+      "Bash(find src/test -name \"*.test.ts\" -exec sed -i '' 's/return this.overrides.get(.insertSemicolons.)/return Promise.resolve(this.overrides.get(''''insertSemicolons''''))/g' {} ;)",
+      "Bash(find src/test -name \"*.test.ts\" -exec sed -i '' 's/return this.overrides.get(.stringQuoteStyle.)/return Promise.resolve(this.overrides.get(''''stringQuoteStyle''''))/g' {} ;)",
+      "Bash(npm test)",
+      "WebSearch",
+      "Bash(npm test -- --grep \"VS Code Organize Imports Behavior\")",
+      "Bash(npm test -- --grep \"preserves blank lines between import groups\")",
+      "Bash(npm test -- --grep \"does NOT automatically create groups based on patterns\")",
+      "Bash(npm test -- --grep \"INVESTIGATION\")",
+      "Bash(npm test -- --grep \"does NOT automatically create groups based on PATTERN MATCHING\")",
+      "Bash(npm test -- --grep \"sorts alphabetically by module path\")",
+      "Bash(npm test -- --grep \"PROOF\")",
+      "Bash(npm run compile-tests)",
+      "Bash(npm test -- --grep \"Edge Cases - Comments and Real-World Scenarios\")",
+      "SlashCommand(/git-quick)"
     ],
     "deny": [],
     "ask": []

CHANGELOG.md

@@ -1,20 +1,9 @@
 # Changelog
 
-## [4.0.0-rc.0] - 2025-10-06
+## [4.0.0-rc.0]
 
-### Release Candidate
+**Initial release** of Mini TypeScript Hero
 
-This is the first release candidate for Mini TypeScript Hero v4.0.0.
-
-**Testing Period:** Please test this RC version and report any issues before the final 4.0.0 release.
-
-**Read the announcement:** [TypeScript Hero is dead (is yet another VS Code extension gone forever?)](https://angular.schule/blog/2025-10-mini-typescript-hero)
-
-## [4.0.0] - TBD
-
-### Added
-
-- **Initial release** of Mini TypeScript Hero
 - Organize imports command (`Ctrl+Alt+O` / `Cmd+Alt+O`)
 - Automatic removal of unused imports
 - Intelligent import sorting (by module path or first specifier)
@@ -32,6 +21,11 @@ This is the first release candidate for Mini TypeScript Hero v4.0.0.
   - Ignored imports (never removed)
   - Disable sorting option
   - Disable removal option
+- **Configuration Priority Order** - Respects `.editorconfig` and VSCode settings before extension settings
+- **Commands:**
+  - `Mini TS Hero: Organize imports` - Sort and remove unused imports (the main command)
+  - `Mini TS Hero: Check for configuration conflicts` - Detect if multiple tools would organize imports
+  - `Mini TS Hero: Toggle legacy mode` - Switch between modern and legacy behavior
 - Modern implementation using ts-morph (v27)
 - Built with esbuild for optimal performance
 

CLAUDE.md

@@ -20,7 +20,7 @@
 ## 🎯 Project Goal
 
 Extract and modernize the "Sort and organize your imports" feature with:
-- ✅ **Comprehensive backward compatibility** with TypeScript Hero settings (verified across 370+ test scenarios)
+- ✅ **Comprehensive backward compatibility** with TypeScript Hero settings
 - ✅ **Modern 2025 tech stack** (ts-morph, esbuild, TypeScript 5.7)
 - ✅ **No dependencies on deprecated libraries** (typescript-parser is 7 years old!)
 - ✅ **Simplified architecture** (no InversifyJS DI container overkill)
@@ -162,7 +162,7 @@ mini-typescript-hero/                     ← Project root
 │   ├── configuration/
 │   │   ├── imports-config.ts             ← 13 config options wrapper
 │   │   └── settings-migration.ts         ← Migrates old TypeScript Hero settings
-│   └── test/                             ← General extension tests (215 tests)
+│   └── test/                             ← General extension tests
 │       ├── imports/import-manager.test.ts
 │       ├── imports/blank-lines.test.ts
 │       └── configuration/settings-migration.test.ts
@@ -171,7 +171,7 @@ mini-typescript-hero/                     ← Project root
 │   ├── old-extension/adapter.ts          ← Adapter for old TypeScript Hero
 │   ├── new-extension/adapter.ts          ← Adapter for new Mini TypeScript Hero
 │   ├── old-typescript-hero/              ← Git submodule (original extension)
-│   └── test-cases/*.test.ts              ← 129 comparison tests
+│   └── test-cases/*.test.ts              ← Comparison tests
 │
 ├── manual-test-cases/                    ← Manual testing scenarios
 ├── package.json                          ← Extension manifest, config schema

CLAUDE_TODO.md

@@ -5051,3 +5051,209 @@ Changes to commit:
 - M src/imports/import-manager.ts (keep LAST default)
 - M src/test/import-manager.test.ts (test 63 updated)
 
+
+---
+
+## Session: 2025-11-07 - VS Code Behavior Documentation & Marketing Clarity
+
+### 1. Current Work Status
+
+#### ✅ Completed Tasks
+
+1. **Created Comprehensive VS Code Behavior Tests**
+   - Added 14 new tests in `src/test/vscode-organize-imports-behavior.test.ts`
+   - Tests execute REAL `editor.action.organizeImports` command (not mocked)
+   - Proves exactly what VS Code can/cannot do with executable evidence
+   - All 309 tests passing
+
+2. **Discovered Critical Timing Requirement**
+   - VS Code's organize imports requires 2-second delay for TypeScript language service
+   - Must wait 1 additional second after command (async execution)
+   - Files can be in os.tmpdir() (VS Code CAN organize standalone files!)
+
+3. **Confirmed #1 Selling Point with 100% Certainty**
+   - VS Code does NOT automatically separate external (node_modules) vs internal (local) imports
+   - Test proves: Mixed `@angular/*` and `./services/*` imports sorted together as ONE group
+   - Mini TypeScript Hero automatically creates blank lines between external/internal groups
+
+4. **Discovered Edge Cases**
+   - Comments WITHIN imports are preserved but treated as group separators
+   - Side-effect imports (`import 'zone.js'`) sorted to END by VS Code
+   - VS Code merges duplicate imports from same module
+
+5. **Cleaned Up Documentation**
+   - README.md: Rewrote "Why Use This Instead of VS Code's Built-in?" section
+   - blog-post.md: Rewrote "Wait, Doesn't VS Code Already Have This?" section
+   - Removed duplicate explanations
+   - Emphasized external/internal grouping as main value (90% use case)
+   - Clarified pattern-based grouping (e.g., `/^@angular/`) is optional enhancement
+
+6. **Updated Audit Documentation**
+   - Added comprehensive section to REQUEST_FOR_AUDIT.md
+   - Documented test strategy, findings, edge cases
+   - Added 6 questions for reviewer feedback
+
+#### 🚫 No In-Progress or Blocked Items
+All tasks completed successfully.
+
+---
+
+### 2. Technical Context
+
+#### Files Created
+
+1. **`src/test/vscode-organize-imports-behavior.test.ts`** (NEW - 309 total tests)
+   - 14 comprehensive tests documenting VS Code's actual behavior
+   - 3 test suites: Group Preservation, Duplicate Merging, Basic Features
+   - 4 tests proving VS Code limitations (our differentiators)
+   - 3 edge case tests (comments, mixed imports, side-effects)
+
+#### Files Modified
+
+1. **`README.md`**
+   - Rewrote "Why Use This Instead of VS Code's Built-in?" section
+   - Removed duplicate explanations about groups
+   - Updated example to show default grouping (`["Modules", "Workspace"]`)
+   - Emphasized automatic external/internal separation as primary value
+
+2. **`blog-post.md`**
+   - Rewrote "Wait, Doesn't VS Code Already Have This?" section
+   - Removed redundant explanations
+   - Added real-world impact example
+   - Emphasized zero-configuration benefit
+
+3. **`REQUEST_FOR_AUDIT.md`**
+   - Added new section: "VS CODE BEHAVIOR DOCUMENTATION TESTS"
+   - Documented test strategy, coverage, findings
+   - Listed interesting edge cases discovered
+   - Added 6 questions for reviewer
+
+4. **`.vscode-test.mjs`**
+   - Briefly added workspace folder to launchArgs (later reverted)
+   - Final version uses default (no workspace needed)
+
+#### Temporary Files (Deleted During Session)
+
+- `src/test/vscode-tmp-test.test.ts` - Created to prove os.tmpdir() works, then deleted
+- `out/test/vscode-tmp-test.test.js` - Compiled version, manually deleted
+
+---
+
+### 3. Important Decisions
+
+#### Architecture Choices
+
+1. **Use Real VS Code APIs for Testing**
+   - Execute actual `editor.action.organizeImports` command
+   - Create real temp files in os.tmpdir()
+   - Use shared test helpers (`createTempDocument`, `deleteTempDocument`)
+   - **Rationale**: Provides executable proof, not speculation
+
+2. **Marketing Messaging: Focus on Default Behavior**
+   - Primary value: Automatic external/internal separation (covers 90% of use cases)
+   - Secondary value: Optional pattern-based grouping (e.g., `/^@angular/`)
+   - **Rationale**: User feedback - "grouping by internal/external usually does the job"
+
+3. **Test Timing Strategy**
+   - 2-second delay before executing organize imports command
+   - 1-second delay after command execution
+   - Total: ~3 seconds per test (acceptable for 14 tests)
+   - **Rationale**: TypeScript language service needs time to analyze file
+
+#### Key Findings
+
+1. **VS Code Does NOT Auto-Group External/Internal**
+   ```typescript
+   // VS Code output (everything mixed):
+   import { Component } from '@angular/core';      // external
+   import { Router } from '@angular/router';        // external
+   import { BookService } from './services/book';  // internal
+   import { UserService } from './services/user';  // internal
+   // NO blank line between external and internal!
+   ```
+
+2. **Comments Treated as Group Separators**
+   - VS Code preserves comments but treats them like blank lines
+   - Creates separate groups around commented imports
+
+3. **Side-Effect Imports Sorted to End**
+   - `import 'zone.js'` moves AFTER named imports
+   - VS Code has special sorting for string-only imports
+
+---
+
+### 4. Next Steps
+
+#### Immediate TODO
+
+1. **Review Test Coverage** (Optional Enhancement)
+   - Consider testing more edge cases based on reviewer feedback
+   - Potential additions: namespace imports, type-only imports, import attributes
+
+2. **Monitor Test Performance**
+   - Current: 14 tests × 3 seconds = ~42 seconds
+   - If test suite grows, consider parallelization or optimization
+
+3. **Update CHANGELOG.md** (Before Release)
+   - Document new VS Code behavior tests
+   - Mention documentation improvements
+
+#### Testing Needed
+
+✅ All tests passing (309 tests, 0 failures)
+✅ ESLint clean
+✅ TypeScript compilation successful
+
+#### Documentation Updates
+
+✅ README.md updated
+✅ blog-post.md updated
+✅ REQUEST_FOR_AUDIT.md updated
+
+**Future Considerations**:
+- Add section to README about side-effect imports behavior?
+- Document comment-as-separator edge case?
+
+---
+
+### 5. Open Questions for Reviewer (from REQUEST_FOR_AUDIT.md)
+
+1. **Test coverage**: Have we covered all relevant edge cases? Any other VS Code behaviors to test?
+
+2. **Marketing claims**: Are our claims in README/blog accurate based on test evidence?
+
+3. **Selling points**: Is "automatic external/internal grouping" our clearest differentiator?
+
+4. **Edge cases**: Should we document the "comments as group separators" behavior in README?
+
+5. **Performance**: 3 seconds per test (14 tests = ~42s) - acceptable or too slow?
+
+6. **Side-effect imports**: Should we test/document how Mini TS Hero handles `import 'zone.js'` vs VS Code?
+
+---
+
+### 6. Test Evidence Summary
+
+**Before This Session**: "VS Code can't automatically create groups" (speculation)
+
+**After This Session**: "VS Code can't automatically create groups" (proven by 14 executable tests)
+
+**Key Test**: `INVESTIGATION: Does VS Code automatically separate external vs internal imports?`
+- Input: Mixed `@angular/*` and `./services/*` imports
+- Output: All sorted together alphabetically (NO blank line between external/internal)
+- Result: `Has blank line between import groups: false`
+
+**Proof Location**: `src/test/vscode-organize-imports-behavior.test.ts:264-283`
+
+---
+
+### 7. Session Metrics
+
+- **Tests Added**: 14 new tests
+- **Total Tests**: 309 passing
+- **Files Modified**: 4
+- **Files Created**: 1
+- **Test Time**: ~60 seconds (full suite)
+- **Session Duration**: ~2 hours
+- **Lines of Test Code**: ~476 lines
+