SeraphimSerapis/tool-eval-bench — GitHub trend

FINDING #1916 · UNIT ID 1213526705

SeraphimSerapis/tool-eval-bench

Tool-calling quality benchmark for LLM serving stacks. 80+ deterministic scenarios testing multi-turn orchestration, safety boundaries, and structured output. Supports vLLM, SGLang, and llama.cpp.

[ PYTHON ][ GITHUB ↗ ]

SURPRISE SCORE

0.00

Score Breakdown

SURPRISE0.0156

ENGAGEMENT0.20

FRESHNESS1.30

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

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

47% OF STARS IN ARCHIVE

Growth Telemetry

VELOCITY /D

3.30

ACCEL

+0.14

RETENTION

0.0%

PEAK 2026-07-15 · FORK-RETENTION 0.0% · 99 STARS / WINDOW

Author Audience

AUDIENCE

172

FOLLOWERS

133

OWNER ★

390

Engagement Signals

FORKS

ISSUE AUTH

PR AUTH

UNIQUE STARGAZERS 99 / 99 (DIVERSITY 1.00)

Why This Is A Finding

SeraphimSerapis/tool-eval-bench собрал 99 звёзд за окно, тогда как у автора всего 133 подписчиков — эффективная аудитория ≈ 172. Это даёт surprise-индекс 0.0156 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 0.0% и 0 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация положительная — рост ещё не выдохся.

Related Findings

RANKS ABOVE 70% OF 6295 FINDINGS

#01shy3130/tickflow-stock-panel[ PYTHON ]0.46 #06Panniantong/Agent-Reach[ PYTHON ]0.29 #10sums001/Windows-Copilot-API[ PYTHON ]0.27 #12HUANGCHIHHUNGLeo/claude-real-video[ PYTHON ]0.24 #14realiti4/claude-swap[ PYTHON ]0.21

METRICS IN CONTEXT

MEDIAN ACROSS ALL 6295 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW

METRICVALUEMEDIANΔ MEDPERCENTILE

SCORE0.010.00+0.00ABOVE 70%

VELOCITY3.304.23-0.93ABOVE 40%

RETENTION0.0%29.4%-29.4 PPABOVE 0%

FORKS2192-71ABOVE 16%

SURPRISE0.020.01+0.00ABOVE 58%