NousResearch/hermes-agent-self-evolution — GitHub trend

◂ BACK TO LEADERBOARD

FINDING #1426 · UNIT ID 1176753219

NousResearch/hermes-agent-self-evolution

⚒ Evolutionary self-improvement for Hermes Agent — optimize skills, prompts, and code using DSPy + GEPA

[ PYTHON ][ ORG ][ GITHUB ↗ ]

SURPRISE SCORE

0.00

Score Breakdown

SURPRISE0.0000485

ENGAGEMENT0.60

FRESHNESS1.25

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

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

11% OF STARS IN ARCHIVE

Growth Telemetry

VELOCITY /D

3.29

ACCEL

-0.05

RETENTION

1.5%

PEAK 2026-04-12 · FORK-RETENTION 60.3% · 592 STARS / WINDOW

Author Audience

AUDIENCE

67,728

FOLLOWERS

8,428

OWNER ★

254,360

Engagement Signals

FORKS

614

ISSUE AUTH

PR AUTH

UNIQUE STARGAZERS 592 / 592 (DIVERSITY 1.00)

Why This Is A Finding

NousResearch/hermes-agent-self-evolution собрал 592 звёзд за окно, тогда как у автора всего 8,428 подписчиков — эффективная аудитория ≈ 67,728. Это даёт surprise-индекс 0.0000485 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 60.3% и 15 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.

Related Findings

RANKS ABOVE 13% OF 1643 FINDINGS

#04cft0808/edict[ PYTHON ]0.0174 #12jundot/omlx[ PYTHON ]0.0123 #16TomBadash/Mouser[ PYTHON ]0.0106 #24dou-jiang/codex-console[ PYTHON ]0.00973 #27wanshuiyin/Auto-claude-code-research-in-sleep[ PYTHON ]0.0095

METRICS IN CONTEXT

MEDIAN ACROSS ALL 1643 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW

METRICVALUEMEDIANΔ MEDPERCENTILE

SCORE0.000.00-0.00ABOVE 13%

VELOCITY3.292.82+0.47ABOVE 58%

RETENTION1.5%4.7%-3.2 PPABOVE 9%

FORKS6141,609-995ABOVE 26%

SURPRISE0.000.00-0.00ABOVE 12%