FINDING #361 · UNIT ID 1300173275
vshulcz/deja-vu
Your agents already solved this. deja finds it — it indexes the sessions your coding agents already wrote to disk, months of history from before you installed it, and recalls them automatically at session start across seventeen harnesses. 84.9% hit@1 on LongMemEval-S, no LLM, no embeddings. One zero-dep binary, fully local.
SURPRISE SCORE
0.00
Score Breakdown
SCORE = SURPRISE × ENGAGEMENT^0.7 × FRESHNESS × VISIBILITY × CONFIDENCE
SURPRISE = WINDOW STARS / DAYS / (AUDIENCE + FLOOR)
16% OF STARS IN ARCHIVE
Growth Telemetry
VELOCITY /D
14.57
ACCEL
-3.25
RETENTION
22.2%
PEAK 2026-08-05 · FORK-RETENTION 20.0% · 102 STARS / WINDOW
Author Audience
AUDIENCE
122
FOLLOWERS
55
OWNER ★
668
Engagement Signals
FORKS
41
ISSUE AUTH
0
PR AUTH
0
UNIQUE STARGAZERS 102 / 102 (DIVERSITY 1.00)
Why This Is A Finding
vshulcz/deja-vu собрал 102 звёзд за окно, тогда как у автора всего 55 подписчиков — эффективная аудитория ≈ 122. Это даёт surprise-индекс 0.0901 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 20.0% и 0 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.
Related Findings
RANKS ABOVE 96% OF 9721 FINDINGS
METRICS IN CONTEXT
MEDIAN ACROSS ALL 9721 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW
METRICVALUEMEDIANΔ MEDPERCENTILE
SCORE0.040.00+0.03ABOVE 96%
VELOCITY14.573.43+11.14ABOVE 85%
RETENTION22.2%32.1%-10.0 PPABOVE 35%
FORKS41130-89ABOVE 25%
SURPRISE0.090.01+0.08ABOVE 95%