- commit
- b53323b684bbb40b915d40c325544501ab3aafe4
- parent
- 8c0e0fcf472c9fbe113aeaac5bd62cb1ce4e70e6
- author
- dami <gazey55@gmail.com>
| M | PLAN.md | 19 | +++++++++++++++++++ |
@@ -758,6 +758,25 @@ a forge repo. The paid Fly `mosaic-nightly` machine is destroyed once
758758 slate's first run has pushed; deploy.md keeps the recipe to recreate it759759 if slate is down. The app machine keeps its emergency field.760760 761+### Found 2026-08-27: the referee is an oracle, and an agent used it762+763+The Codex agent's script commits every unknown cell of a revealed tile764+as 0 in one batch, reads the rejections (which echo the value), commits765+the complements, and never deduces anything: 723 tiles, 41k accepted766+cells, 20,557 wrong, top of the day's leaderboard. Within the stated767+rules — the tool text itself says a wrong binary value is full768+information — and exactly what a per-cell verdict invites. A Takuzu is769+self-verifying by its rules; the referee must not be a "check this770+cell" button.771+772+Decision to make (not shipped yet, it changes the live rules): commits773+are claims, verified per tile when the agent claims it complete; a774+failed claim resets the tile's claimed cells, costs reward (at least775+the tile's cells) and a cooldown, and reveals only "wrong", not where;776+wrong% on the leaderboard; the per-cell verdict survives only as a777+localhost practice knob. A genuine m-way ambiguity then costs at most778+m-1 failed claims, which is the price of a guess made real.779+761780 ### Status of the v2 items under this goal762781 763782 1 real packed run: done (v2 packings and v3 scatter fields played, logs