claude
787cc56522
eval: add the held-out ecosystem reach fixture (V-405)
...
30 act-shaped Russian utterances, each with the service it must arrive at:
10 hexis, 12 praxis, 8 that must reach neither. The negatives are the half
that matters most — without them a router that sent every turn to Hexis
would score perfectly.
want_capability records which Praxis arm the fn should land on. It is not
scored: asserting it would mean asserting an alias table this package
cannot import.
2026-08-04 06:21:40 +04:00
..
2026-08-01 14:36:17 +04:00
2026-08-04 03:08:09 +04:00
2026-08-04 05:23:11 +04:00
2026-08-02 02:28:53 +04:00
2026-08-04 05:15:02 +04:00
2026-08-01 14:33:26 +04:00
2026-08-02 09:11:11 +04:00
2026-08-04 03:33:19 +04:00
2026-08-01 14:08:32 +04:00
2026-08-02 09:11:11 +04:00
2026-08-04 05:46:15 +04:00
2026-08-02 03:28:49 +04:00
2026-08-03 12:27:04 +04:00
2026-08-04 05:15:01 +04:00
2026-08-01 14:11:57 +04:00
2026-08-01 14:36:17 +04:00
2026-08-01 13:57:26 +04:00
2026-08-04 03:18:42 +04:00
2026-08-04 03:17:29 +04:00
2026-08-02 15:08:07 +02:00
2026-08-01 14:01:01 +04:00
2026-08-04 03:00:11 +04:00
2026-07-31 16:07:35 +04:00
2026-08-04 05:15:01 +04:00
2026-08-04 06:21:40 +04:00
2026-07-31 02:45:30 +04:00
2026-08-01 14:21:36 +04:00
2026-08-01 14:06:38 +04:00
2026-08-01 14:12:43 +04:00
2026-08-04 05:39:52 +04:00
2026-08-02 03:28:49 +04:00
2026-08-01 14:16:56 +04:00
2026-08-02 03:28:49 +04:00
2026-08-02 03:28:49 +04:00
2026-07-31 10:09:32 +04:00
2026-08-01 14:06:00 +04:00
2026-08-01 14:21:46 +04:00
2026-08-01 20:05:52 +04:00
2026-08-04 03:25:42 +04:00
2026-08-01 14:05:13 +04:00
2026-08-01 04:22:52 +04:00
2026-08-02 02:37:45 +04:00
2026-08-02 15:08:07 +02:00
2026-08-01 14:16:56 +04:00