Level 1 tests validate each vector store's search() returns similarity
scores (higher = better) with correct ordering using known vectors.
Level 2 tests validate the full Memory.add() + Memory.search(threshold=...)
pipeline end-to-end. Both levels are standalone scripts run via CLI args
(e.g., `python test_score_normalization.py faiss`), not collected by pytest.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The threshold parameter in Memory.search() was silently broken for 11
vector stores because they returned raw distance scores (lower = better)
while the threshold check assumed similarity scores (higher = better).
This caused threshold filtering to be inverted — good matches got
dropped and bad matches passed through.
Convert all affected vector stores to return similarity scores:
- Cosine distance stores: score = max(0.0, 1.0 - distance)
- L2 distance stores: score = 1.0 / (1.0 + distance)
- Stores computing similarity then discarding it: return similarity directly
Fixes#4453
Related: #3283
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
