kind string | loop string | improve_skill_version string | iteration int64 | road int64 | remaining int64 | memory_lines int64 | refused int64 | skill dict | policy string | headroom dict | last dict | derived_by string | improvements_applied list | archived_versions int64 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
sagi.skill_publication/1 | jaimla | 0.0.4 | 4 | 1,000 | 996 | 4 | 0 | {
"source": "PYTHAI/jaimla-loop SKILL.md",
"version": "0.0.4",
"sha256": "d8d92f007072d59cb91082ac79b6d13f968ec5b7bf68d1b6604b10831b3aea91",
"bytes": 1178
} | fully automatic (operator, 2026-09-26): each verified iteration publishes the next skill | {
"last": 135,
"mean_last_20": 124.8
} | {
"n": 4,
"version": "0.0.4",
"date": "2026-09-26T00:00:00",
"limit": 160,
"completion_tokens": 25,
"headroom": 135
} | scripts/savante_loop.py with the canon's bind/skill_history.py | [] | 4 |
No dataset card yet
- Downloads last month
- -