Phase 14: LLM-grounded encounter-balancing tip

- src/lib/assistant/prompts.ts: balanceSuggestionSchema + buildBalancePrompt
  (system message leads with the grounding constraint; restricts choices to the
  provided compendium candidates).
- useEncounterAdvisor hook: builds context → picks system-correct candidates →
  asks the LLM (when configured) for a structured pick, validates + filters names
  to the candidate set, else falls back to the deterministic buildSuggestedEncounter.
  Apply adds the creatures transactionally via encountersRepo.mutate.
- EncounterTipCard in the combat tracker: leads with the detected difficulty
  tendency (or the current difficulty), offers a one-click grounded suggestion
  (AI or deterministic) with propose→confirm Apply.

prompts unit tests + e2e for both the deterministic apply flow and the AI path
(stubbed provider); the AI test also confirms candidate-grounding (only
shortlisted creatures survive).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

src/features/combat/EncounterTracker.tsx

@@ -27,6 +27,7 @@ import { cn } from '@/lib/cn';
 import { Button } from '@/components/ui/Button';
 import { Input, Select } from '@/components/ui/Input';
 import { NumberField } from '@/components/ui/NumberField';
+import { EncounterTipCard } from '@/features/assistant/EncounterTipCard';
 
 export function EncounterTracker({ encounter, campaign }: { encounter: Encounter; campaign: Campaign }) {
   const characters = useCharacters(campaign.id);
@@ -129,6 +130,8 @@ export function EncounterTracker({ encounter, campaign }: { encounter: Encounter
         </div>
       </div>
 
+      <EncounterTipCard campaign={campaign} encounter={encounter} />
+
       <AddCombatantBar encounter={encounter} characters={characters} onAdd={(c) => mutate((e) => addCombatant(e, c))} />
 
       {/* Combatant list */}