Test Memory System

cd resources/local-brain-search

# Get current Q-value for a test note
cat data/q_values.json | python -c "import json,sys; d=json.load(sys.stdin); print(d.get('02-Permanent/Dopamine.md', 'not found'))"

# Log a read event
./run_learning.sh log read "02-Permanent/Dopamine.md" --query "test"

# Check Q-value increased
cat data/q_values.json | python -c "import json,sys; d=json.load(sys.stdin); print(d.get('02-Permanent/Dopamine.md', 'not found'))"

cd resources/local-brain-search

# First, artificially boost a note's Q-value
python -c "
import json
with open('data/q_values.json', 'r') as f:
    q = json.load(f)
q['02-Permanent/Identity.md'] = 1.5  # High Q-value
with open('data/q_values.json', 'w') as f:
    json.dump(q, f, indent=2)
print('Q-value set for Identity.md')
"

# Search for something where Identity.md is somewhat relevant
./run_search.sh "belief systems self" --mode spreading --limit 10 --json

# Check if Identity.md ranks higher due to Q-value boost

cd resources/local-brain-search

# Count events before
BEFORE=$(wc -l < data/usage_history.jsonl)

# Run search with --no-track
./run_search.sh "no track test" --mode static --no-track

# Count events after
AFTER=$(wc -l < data/usage_history.jsonl)

echo "Events should be same: before=$BEFORE, after=$AFTER"

cd resources/local-brain-search

# Print current configuration
python memory_config.py

# Verify spreading uses config
python -c "
from memory_config import MEMORY_CONFIG
print('Spreading max_iterations:', MEMORY_CONFIG['spreading']['max_iterations'])
print('Learning enabled:', MEMORY_CONFIG['learning']['enabled'])
print('Q-weight:', MEMORY_CONFIG['learning']['q_weight'])
"

cd resources/local-brain-search

# Benchmark static search
time (for i in {1..5}; do ./run_search.sh "dopamine" --mode static --limit 10 --no-track > /dev/null; done)

# Benchmark spreading search
time (for i in {1..5}; do ./run_search.sh "dopamine" --mode spreading --limit 10 --no-track > /dev/null; done)

cd resources/local-brain-search

# Empty query
./run_search.sh "" --mode spreading --limit 5 2>&1

# Very long query
./run_search.sh "$(python -c 'print("dopamine " * 100)')" --mode spreading --limit 5 2>&1

# Non-existent topic
./run_search.sh "xyznonexistenttopicxyz" --mode spreading --limit 5 --json

# Unicode query
./run_search.sh "意識 consciousness" --mode spreading --limit 5 --json

REPORT_DIR="resources/memory-test-reports"
REPORT_FILE="$REPORT_DIR/test-report-$(date +%Y-%m-%d-%H%M).md"
mkdir -p "$REPORT_DIR"

# Memory System Test Report

**Date:** YYYY-MM-DD HH:MM
**Tester:** [name]
**System Version:** Cornelius v01.25

## Summary

| Component | Status | Notes |
|-----------|--------|-------|
| Intent Classification | ✓/✗ | |
| Spreading Activation | ✓/✗ | |
| Lateral Inhibition | ✓/✗ | |
| Usage Tracking | ✓/✗ | |
| Q-Value Learning | ✓/✗ | |
| Configuration Loading | ✓/✗ | |
| Performance | ✓/✗ | |

## Detailed Results

### Intent Classification
- Accuracy: X%
- Failures: [list]

### Spreading vs Static
- Result overlap: X/5
- Spreading found cross-domain connections: Y/N
- Example good result: [describe]

### Lateral Inhibition
- Hub dominance reduced: Y/N
- Diversity improved: Y/N

### Learning System
- Events tracked: X
- Q-values updated correctly: Y/N
- Ranking adjustment working: Y/N

### Performance
- Static search avg: Xms
- Spreading search avg: Xms
- Acceptable: Y/N

## Issues Found

1. [Issue description]
   - Severity: High/Medium/Low
   - Steps to reproduce: [steps]
   - Suggested fix: [suggestion]

## Recommendations

1. [Recommendation]
2. [Recommendation]

## Next Steps

- [ ] Address issues found
- [ ] Move to Phase 2 (Extended Graph) if all passes
- [ ] Enable spreading as default mode if stable