- Simplified loadPersistedVocabularyData() to use only VocabularyProgressManager - Updated calculateVocabularyProgress() to use unified data structure - Removed old system references from knowledge panel data loading - Fixed field names (drsDiscovered, drsMastered) for unified system - Knowledge panel now displays vocabulary progress correctly ✅ TESTED: Vocabulary Knowledge panel working with unified system 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com> |
||
|---|---|---|
| .. | ||
| ai-validation | ||
| root-cleanup | ||
| README.md | ||
Tests Directory
Structure
ai-validation/
Collection of AI scoring system validation tests created during development.
Key Test Files:
test-all-outputs.js- Comprehensive test of all exercise types with both providerstest-variance-quick.js- Quick variance test without cache interferencetest-spanish-bug.js- Specific test for translation bug that was fixedtest-strict-scoring.js- Validation of strict scoring rulestest-final-validation.js- Production readiness validation
Debug Files:
debug-*.js- Various debugging utilities for AI system developmentcompare-instances.js- Comparison between LLMValidator and direct IAEngine
Historical Tests:
- Tests documenting the development process and bug fixes
- Comprehensive validation achieving 100% success rate
Running Tests
From project root:
# Run comprehensive validation
node tests/ai-validation/test-all-outputs.js
# Quick variance test
node tests/ai-validation/test-variance-quick.js
# Test specific scenarios
node tests/ai-validation/test-strict-scoring.js
Notes
- All tests require proper API keys in
.envfile - Cache is currently disabled for accurate testing
- Tests validate real AI responses (no mock/fallback)
- Achieved 100% validation success rate in December 2024