feat: add CI workflow, lifecycle integration tests, and mark completed plan items

- Add GitHub Actions CI: test, typecheck, lint on push/PR to main
- Add fight lifecycle integration tests: full pipeline, ELO advantage,
  combo scaling, type exhaustion, answer verification (7 tests)
- Kaplay already lazy-loaded via route-level code splitting
- Total: 135 tests across 7 test files

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

server/src/engine/lifecycle.test.ts

@@ -0,0 +1,142 @@
+/**
+ * End-to-end lifecycle tests for the fight scoring pipeline.
+ * Tests the full challenge → response → scoring → elo → tier flow
+ * without requiring database access.
+ */
+import { describe, it, expect } from 'vitest'
+import { pickChallenge, getAllChallengeTypes, type Challenge } from './challenges.js'
+import { checkAnswer } from './answers.js'
+import { scoreRound, calculateElo, calculateTier } from './scoring.js'
+import { mockResponse } from './mock.js'
+
+function simulateFight(eloA: number, eloB: number, rounds = 5) {
+  const botA = { id: 'a1', name: 'FighterA' }
+  const botB = { id: 'b1', name: 'FighterB' }
+  const usedTypes = new Set<string>()
+  let comboA = 0
+  let comboB = 0
+  const results = []
+
+  for (let i = 0; i < rounds; i++) {
+    const challenge = pickChallenge(usedTypes, null)
+    usedTypes.add(challenge.type)
+
+    const respA = mockResponse(challenge, 'confident', eloA)
+    const respB = mockResponse(challenge, 'clueless', eloB)
+
+    const result = scoreRound(
+      challenge, botA, botB,
+      { answer: respA.answer, timeMs: respA.timeMs, timedOut: respA.timedOut, error: respA.error },
+      { answer: respB.answer, timeMs: respB.timeMs, timedOut: respB.timedOut, error: respB.error },
+      null, comboA, comboB,
+    )
+
+    if (result.winnerId === botA.id) { comboA++; comboB = 0 }
+    else if (result.winnerId === botB.id) { comboB++; comboA = 0 }
+
+    results.push(result)
+  }
+
+  return results
+}
+
+describe('fight lifecycle', () => {
+  it('full pipeline: challenge → mock response → scoring → results', () => {
+    const results = simulateFight(1800, 1000)
+
+    expect(results.length).toBe(5)
+    for (const r of results) {
+      expect(r.botAScore).toBeGreaterThanOrEqual(0)
+      expect(r.botBScore).toBeGreaterThanOrEqual(0)
+      expect(r.botADamage).toBeGreaterThanOrEqual(0)
+      expect(r.botBDamage).toBeGreaterThanOrEqual(0)
+      expect(typeof r.narration).toBe('string')
+      expect(typeof r.isCritical).toBe('boolean')
+    }
+  })
+
+  it('higher ELO bot wins more rounds on average', () => {
+    let highWins = 0
+    let lowWins = 0
+    // Run many fights to average out randomness
+    for (let f = 0; f < 20; f++) {
+      const results = simulateFight(1800, 900)
+      for (const r of results) {
+        if (r.winnerId === 'a1') highWins++
+        else if (r.winnerId === 'b1') lowWins++
+      }
+    }
+    expect(highWins).toBeGreaterThan(lowWins)
+  })
+
+  it('Elo updates reflect fight outcome', () => {
+    const eloA = 1400
+    const eloB = 1400
+    const { newWinnerElo, newLoserElo } = calculateElo(eloA, eloB)
+    expect(newWinnerElo).toBeGreaterThan(eloA)
+    expect(newLoserElo).toBeLessThan(eloB)
+    // Sum should be roughly preserved (zero-sum)
+    expect(Math.abs((newWinnerElo + newLoserElo) - (eloA + eloB))).toBeLessThan(1)
+  })
+
+  it('tier progresses with wins and Elo', () => {
+    expect(calculateTier(1200, 0)).toBe(0)  // no wins
+    expect(calculateTier(1200, 1)).toBe(1)  // bronze
+    expect(calculateTier(1200, 3)).toBe(2)  // silver
+    expect(calculateTier(1350, 7)).toBe(3)  // gold
+    expect(calculateTier(1500, 15)).toBe(4) // platinum
+    expect(calculateTier(1700, 25)).toBe(5) // diamond
+    expect(calculateTier(1900, 40)).toBe(6) // legend
+  })
+
+  it('no type repeats in single fight until exhausted', () => {
+    const usedTypes = new Set<string>()
+    const allTypes = getAllChallengeTypes()
+
+    // Pick challenges for all 16 types — no repeats
+    for (let i = 0; i < allTypes.length; i++) {
+      const c = pickChallenge(usedTypes, null)
+      expect(usedTypes.has(c.type)).toBe(false)
+      usedTypes.add(c.type)
+    }
+    expect(usedTypes.size).toBe(allTypes.length)
+
+    // After exhausting all types, reset works
+    const c = pickChallenge(usedTypes, null)
+    expect(c).toBeTruthy()
+  })
+
+  it('checkAnswer integrates with challenge answers', () => {
+    // Pick factual challenges and verify correct answers score 1.0
+    for (let i = 0; i < 50; i++) {
+      const c = pickChallenge(new Set(), null)
+      if (c.scoring === 'factual' && c.answers && c.answers.length > 0) {
+        const score = checkAnswer(c.answers[0], c.answers)
+        expect(score).toBe(1.0)
+      }
+    }
+  })
+
+  it('combo buildup increases damage across rounds', () => {
+    const challenge: Challenge = {
+      type: 'riddle',
+      label: 'Test',
+      prompt: 'Q?',
+      answers: ['4'],
+      scoring: 'factual',
+      baseDamage: 20,
+      timeout_ms: 8000,
+    }
+    const botA = { id: 'a1', name: 'A' }
+    const botB = { id: 'b1', name: 'B' }
+    const resp = { answer: '4', timeMs: 200, timedOut: false, error: false }
+    const wrong = { answer: 'x', timeMs: 200, timedOut: false, error: false }
+
+    const r0 = scoreRound(challenge, botA, botB, resp, wrong, null, 0, 0)
+    const r3 = scoreRound(challenge, botA, botB, resp, wrong, null, 3, 0)
+    const r5 = scoreRound(challenge, botA, botB, resp, wrong, null, 5, 0)
+
+    expect(r3.botADamage).toBeGreaterThan(r0.botADamage)
+    expect(r5.botADamage).toBeGreaterThan(r3.botADamage)
+  })
+})