Merge pull request 'Reduce turn latency: cap recent events (10→5) + concise-reasoning constraint (v0.9.0)' (#23) from feat/turn-latency-hardening into master

CHANGELOG.md

@@ -2,6 +2,9 @@
 
 All notable changes to **History** are listed here, newest first.
 
+## v0.9.0 — 2026-10-11 20:57
+- **Reduced turn latency (prompt-level):** capped `recentEvents` in the turn prompt from the last 10 to the last 5 events, and added a HARD CONSTRAINT asking the reasoning model to keep its internal analysis brief before emitting the JSON. This shortens the chain-of-thought (the 170–223s turn driver) and lowers the risk of exhausting the 4000-token output budget before the answer (the original 502 failure mode). Reasoning stays enabled (`reasoning_effort: low`); the deterministic stub, the `TurnResult` contract, `max_tokens`, and `reasoning_effort` are all unchanged.
+
 ## v0.8.0 — 2026-10-11 20:29
 
 - **Remove a saved game (#20)** — every saved game now has a delete control: a trash button on its card in the home screen's "Saved games" section opens a confirmation dialog ("Delete this game?" showing the scenario/turn/date) with Cancel and a destructive Delete. The new `DELETE /api/games/:id` route removes the game's row from SQLite and returns `{ deleted: true }` (404 if the id is unknown).

package.json

@@ -1,6 +1,6 @@
 {
   "name": "history",
-  "version": "0.8.0",
+  "version": "0.9.0",
   "private": true,
   "type": "module",
   "description": "AI-driven alternate-history simulator — the LLM is the world engine (Pax Historia style).",

server/engine.test.ts

@@ -1,10 +1,11 @@
 import { describe, expect, it } from 'vitest';
 import type { GameState, Preset, RoundItem } from '../shared/types';
-import type { TurnResult } from '../shared/protocol';
+import { makeEvent, type TurnResult } from '../shared/protocol';
 import type { LlmClient } from './llm';
 import { makeStubLlmClient } from './stub';
 import {
   applyTurnResult,
+  buildSystemPrompt,
   buildUserPrompt,
   isMeaningful,
   parseTurnResult,
@@ -279,6 +280,44 @@ describe('buildUserPrompt (batched round)', () => {
     expect(parsed.roundInstruction).toMatch(/no action/i);
     expect(parsed.roundInstruction).toMatch(/own momentum/i);
   });
+
+  it('caps recentEvents at the last 5 (in order) when the state has more than 5 events', () => {
+    const state = baseState();
+    const events = Array.from({ length: 8 }, (_, i) =>
+      makeEvent(
+        {
+          nation: 'uk',
+          type: 'political',
+          title: `Event ${i + 1}`,
+          text: `Narrative ${i + 1}`,
+          outcome: 'neutral',
+        },
+        i + 1,
+        `1939-09-0${i + 1}`,
+      ),
+    );
+    state.events = events;
+
+    const parsed = JSON.parse(buildUserPrompt(state, [])) as {
+      state: { recentEvents: Array<{ title: string; turn: number; date: string }> };
+    };
+
+    expect(parsed.state.recentEvents).toHaveLength(5);
+    // the last 5 of the 8 input events, in order
+    expect(parsed.state.recentEvents[0].title).toBe(events[3].title); // 'Event 4'
+    expect(parsed.state.recentEvents[4].title).toBe(events[7].title); // 'Event 8'
+    expect(parsed.state.recentEvents.map((e) => e.title)).toEqual([
+      events[3].title,
+      events[4].title,
+      events[5].title,
+      events[6].title,
+      events[7].title,
+    ]);
+  });
+
+  it('asks the model to keep its internal reasoning brief (concise-CoT constraint)', () => {
+    expect(buildSystemPrompt(preset)).toMatch(/reasoning brief|a few sentences/i);
+  });
 });
 
 describe('queueRoundItem', () => {

server/engine.ts

@@ -59,6 +59,7 @@ export function buildSystemPrompt(preset: Preset): string {
     'HARD CONSTRAINTS',
     '- Use only nation ids and ISO_A3 codes that exist in the provided state.',
     '- Prefer 1-4 events per turn; keep each text 1-3 sentences.',
+    '- Keep your internal reasoning brief (a few sentences) and go straight to the JSON result; do not over-deliberate or re-derive facts already present in the state.',
     '- Keep every stat value between 0 and 100.',
     '- Only change borders when war or a decisive political shift justifies it.',
     '- If a field is not relevant this turn, set it to an empty array.',
@@ -95,7 +96,7 @@ export function buildUserPrompt(state: GameState, items: RoundItem[]): string {
       nations: Object.fromEntries(
         Object.entries(state.nations).map(([id, n]) => [id, nationForPrompt(n)]),
       ),
-      recentEvents: state.events.slice(-10).map((e) => ({
+      recentEvents: state.events.slice(-5).map((e) => ({
         turn: e.turn,
         date: e.date,
         nation: e.nation,