Found 2 medium-priority issues and 1 low. (1) Resting↔philosophizing rapid cycling wastes LLM calls because there’s no `recentlyLeftState` cooldown on the philosophizing FSM transition — add 60-second re-entry guard. (2) Hunting has no failure escalation, allowing 7+ consecutive misses — add +10% success bonus per consecutive fail (capped at +40%). (3) Duplicate atmosphere_shift narration text broadcast to clients — add dedup check against recent history.
(MEDIUM): Resting↔Philosophizing Rapid Cycling Wastes LLM Calls
Fix: Add a `recentlyLeftState` guard to both philosophizing transitions in `creature-fsm.ts`: (code change) And the same for the EXPLORING/WANDERING→PHILOSOPHIZING transition at line 1435:
(code change) This enforces a 60-second cooldown between philosophizing sessions, preventing the rapid cycle while…
(MEDIUM): Hunting Has No Failure Escalation — 7 Consecutive Failures
Fix: Add a per-creature consecutive failure bonus in `resource-system.ts`. Track failures and boost success chance by +10% per consecutive miss, capping at +40%: (code change) The counter should be stored on the pet-brain or creature traits, not passed in as a trait. The implementation detail: add…
(LOW): Duplicate Atmosphere Shift Narration Text
Fix: Add a last-narration dedup check in `world-brain.ts` after event creation (around line 247): (code change) —
(LOW): Zhraa Social Think JSON Parse Failures (2x)
Fix applied via code changes.
Files Modified
creature-fsm.tsresource-system.tsworld-brain.ts