BitBrain gun (id 16): Pattern-relative ADE+SBC aim corrector, default off

Wire the verified common_libs/bitbrain ADE+SBC library into ModularBot as a
fine-grained angular corrector on top of Pattern's prediction, the shape the
offline gate test measured (argmax readout over N correction classes).

- common_libs/guns/bitbrain_gun.nim: new gun. Input = the existing TMHorizon
  53 bits (tmhBaseBits + tmhLits); output = argmax class centre over
  +-TR_BITBRAIN_RANGE, applied by rotating the Pattern point around the shooter
  exactly as tmhApplyShift does. Label = the +h-tick fact from TmHorizonGun's
  own observation ring (never across a round). Prequential (defer + resolve).
  AD layer synthesised online for our binary inputs (center=0): heuristic
  cold-start thresholds + running-histogram ~1% percentile init + the library's
  adaptThresholds. Memory modes perRound (default, measured best) / retained /
  decay (periodic partial SBC wipe). Lazy network build + local RNG, so the
  default path builds nothing and consumes no global randomness.
- tm_horizon.nim: export tmhUpdateHistory and add tmhObservedAt (label seam).
- selector.nim: register BITBRAIN at rack id 16, default rmOff, in the SAME
  commit as the id and the wiring (the aed579b admission bug is not repeated).
- ModularBot.nim: id 16 wired through predict/spawn/onResult/resets/colors,
  arrays grown 16->17, spawn gated on rack admission, per-round/per-battle/
  target reset hooks.
- env_report.nim: report every TR_BITBRAIN_* knob + add names to the known set.
- tests: update the rack length literals; new test_bitbrain_registration
  (default-parity: off, lazy, global-RNG clean).

Guard counts unchanged: rack 48, tm_pattern_registration 20, vbullet_admit 12,
env_report 25, and the rest of the suite green.

common_libs/guns/tm_horizon.nim

@@ -546,6 +546,19 @@ proc ringValidCount*(g: TmHorizonGun): int =
   for i in 0..<TMH_POS_RING:
     if g.ringValid[i]: inc result
 
+proc tmhObservedAt*(g: TmHorizonGun, tick: int, selfX, selfY: float
+                   ): tuple[ok: bool, bearing, dist: float, lastSeenTick: int] =
+  ## Exported seam for the BitBrain corrector: the enemy's OBSERVED bearing and
+  ## distance (relative to `(selfX, selfY)`) at `tick`, if the observation ring
+  ## holds that exact tick. `lastSeenTick` lets the caller apply the same
+  ## staleness rule TMHorizon uses before trusting the label. `ok = false` when
+  ## the tick is missing (e.g. across a round boundary or before warm-up).
+  let r = g.ringAt(tick)
+  if not r.ok:
+    return (false, 0.0, 0.0, 0)
+  (true, arctan2(r.t.ey - selfY, r.t.ex - selfX),
+   hypot(r.t.ex - selfX, r.t.ey - selfY), r.t.lastSeenTick)
+
 proc ringHas*(g: TmHorizonGun, tick: int): bool =
   ## Observability / test seam: is there a valid observation exactly at `tick`?
   ## Used to prove a round boundary drops the old positions (so a label can never
@@ -598,7 +611,7 @@ proc bulletFeatures(g: var TmHorizonGun, state: WorldState): tuple[tta: int, lat
       result.lat = b.ux * vy - b.uy * vx
   g.bulletCount = w
 
-proc tmhUpdateHistory(g: var TmHorizonGun, state: WorldState) =
+proc tmhUpdateHistory*(g: var TmHorizonGun, state: WorldState) =
   ## Once per tick: write the observation ring, advance the reversal counter and
   ## the self-energy-drop bullet proxy.
   let slot = ((state.tick mod TMH_POS_RING) + TMH_POS_RING) mod TMH_POS_RING