ai-boost/awesome-harness-engineering — GitHub trend

FINDING #2795 · UNIT ID 1195371847

ai-boost/awesome-harness-engineering

Awesome list for AI agent harness engineering: tools, patterns, evals, memory, MCP, permissions, observability, and orchestration.

[ PYTHON ][ GITHUB ↗ ]

SURPRISE SCORE

0.00

Score Breakdown

SURPRISE0.0115

ENGAGEMENT0.42

FRESHNESS1.00

SCORE = SURPRISE × ENGAGEMENT^0.7 × FRESHNESS × VISIBILITY × CONFIDENCE

SURPRISE = WINDOW STARS / DAYS / (AUDIENCE + FLOOR)

5% OF STARS IN ARCHIVE

Growth Telemetry

VELOCITY /D

27.57

ACCEL

-0.43

RETENTION

57.1%

PEAK 2026-08-25 · FORK-RETENTION 65.4% · 193 STARS / WINDOW

Author Audience

AUDIENCE

2,360

FOLLOWERS

676

OWNER ★

16,840

Engagement Signals

FORKS

473

ISSUE AUTH

PR AUTH

UNIQUE STARGAZERS 193 / 193 (DIVERSITY 1.00)

Why This Is A Finding

ai-boost/awesome-harness-engineering собрал 193 звёзд за окно, тогда как у автора всего 676 подписчиков — эффективная аудитория ≈ 2,360. Это даёт surprise-индекс 0.0115 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 65.4% и 0 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.

Related Findings

RANKS ABOVE 71% OF 9771 FINDINGS

#05MddIdd/mdd-sim-gateway[ PYTHON ]0.41 #08zeenie-ai/OpenCompany[ PYTHON ]0.38 #15litaolemo/xtquant_big_convert[ PYTHON ]0.33 #23917Dhj/DeepPaperNote[ PYTHON ]0.27 #24LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler[ PYTHON ]0.27

METRICS IN CONTEXT

MEDIAN ACROSS ALL 9771 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW

METRICVALUEMEDIANΔ MEDPERCENTILE

SCORE0.010.00+0.00ABOVE 71%

VELOCITY27.573.71+23.86ABOVE 92%

RETENTION57.1%39.3%+17.8 PPABOVE 77%

FORKS47391+382ABOVE 85%

SURPRISE0.010.01+0.00ABOVE 56%