Fix 1-step termination bug, fix done field in trajectory, recalibrate seeds

server/constants.py

@@ -46,7 +46,7 @@ HOSPITAL_BREACH_POINT = 0.00    # At or below this → hospital has collapsed
 SPREAD_RATE_MIN       = 0.04    # Slowest possible true spread rate per day
 SPREAD_RATE_MAX       = 0.15    # Fastest possible true spread rate per day
 GROWTH_HINT_NOISE     = 0.03    # Random noise added to growth_rate_hint (± value)
-TREATMENT_REDUCTION   = 0.10    # Allocating also reduces existing infection
+TREATMENT_REDUCTION   = 0.06    # Allocating also reduces existing infection
 ALLOCATE_REDUCTION    = 0.10    # How much one 'allocate' reduces spread this step
 RESTRICT_REDUCTION    = 0.05    # How much one 'restrict' reduces spread per step
 SPILLOVER_RATE        = 0.01    # Fraction of infection that spreads to adjacent districts per day

server/environment.py

@@ -176,7 +176,7 @@ class EpidemicContainmentEnv(Environment):
                 "total_steps":         result.total_steps,
                 "task_name":           self._task_name,
             }
-
+            self._trajectory[-1].done = True
         # ── 11. Build and return observation ──────────────────────────────────
         final_message = terminal_message if terminal_message else message
 

server/tasks/task_hard.py

@@ -29,7 +29,7 @@ class HardTask(BaseTask):
         # All districts start with small infections that grow at different rates.
         # The 3-day lag means the agent won't see today's true rates until day 3.
         # Districts with high true_spread_rate will accelerate invisibly.
-        seed_infections = [0.07, 0.05, 0.08, 0.04, 0.10, 0.06]
+        seed_infections = [0.12, 0.08, 0.14, 0.06, 0.16, 0.09]
 
         # Pre-populate infection_history with 3 days of identical
         # starting values so the lag mechanic works from step 1.

server/tasks/task_medium.py

@@ -28,7 +28,7 @@ class MediumTask(BaseTask):
         # Districts 0 and 2 are seeded with outbreaks (non-adjacent).
         # Districts 1 and 3 are clean but will receive spillover.
         # Agent must choose which outbreak to tackle first.
-        seed_infections = [0.12, 0.03, 0.10, 0.03]
+        seed_infections = [0.25, 0.06, 0.22, 0.06]
 
         return CityState(
             day                 = 0,