Best use case
game-theory-debate is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
博弈论决策引擎：当用户请求决策时，自动启动多策略博弈分析，找到优势策略，均衡思维，接受失败作为学习代价。
Teams using game-theory-debate should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/game-theory-debate/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How game-theory-debate Compares
| Feature / Agent | game-theory-debate | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
博弈论决策引擎：当用户请求决策时，自动启动多策略博弈分析，找到优势策略，均衡思维，接受失败作为学习代价。
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
Related Guides
AI Agents for Marketing
Discover AI agents for marketing workflows, from SEO and content production to campaign research, outreach, and analytics.
AI Agents for Startups
Explore AI agent skills for startup validation, product research, growth experiments, documentation, and fast execution with small teams.
AI Agents for Coding
Browse AI agent skills for coding, debugging, testing, refactoring, code review, and developer workflows across Claude, Cursor, and Codex.
SKILL.md Source
# 🎮 Game Theory Debate（博弈论决策引擎）
> 每一个决策都是一场博弈。你的选择、我的建议、环境的约束——都在payoff矩阵里。
---
## 🔄 自动触发机制
### 触发词（任一出现即启动）
| 类型 | 示例 |
|------|------|
| **决定请求** | 「你觉得哪个好」「帮我选」「听你的」 |
| **判断请求** | 「应该怎么做」「这样做对不对」「值不值得」 |
| **方案对比** | 「A还是B」「方案1vs方案2」「哪个更优」 |
| **犹豫表达** | 「我在纠结」「拿不定主意」「不知道怎么选」 |
| **反思请求** | 「复盘一下」「哪里做错了」「下次怎么改」 |
### 不触发
- 闲聊、寒暄
- 纯执行指令（步骤已明确）
- 简单事实查询
- 情绪安抚（除非涉及决策）
---
## ⚙️ 工作流程
```
用户决策请求
│
▼
[1] 裁判模式启动
- 判断博弈类型（零和/非零和/混合）
- 选择参与者策略
│
▼
[2] 多策略出牌（2-4轮）
- 各方提交核心主张 + payoff评估
- 交叉批评与防守
- 寻找Nash均衡
│
▼
[3] 均衡分析
- 淘汰劣势策略
- 识别focal point（直觉汇聚）
- 输出推荐策略
│
▼
[4] 决策输出
- 推荐方案（带置信度）
- 备选方案
- 风险提示
│
▼
[5] 结果归档（事后可选）
- 如果用户反馈结果
- 记录决策效果
- 更新策略库
```
---
## 🎭 参与者策略库
每个博弈至少有 2-3 个策略视角：
| 策略 | 特点 | 代表声音 |
|------|------|---------|
| **激进派** | 最大化收益，愿意冒险 | 「搏一搏，单车变摩托」 |
| **保守派** | 最小化风险，确定性优先 | 「稳字当头」 |
| **经济学家** | 成本收益分析 | 「投入产出比」 |
| **工程师** | 可行性第一 | 「能不能做出来」 |
| **用户代言人** | 从老豆视角出发 | 「老豆的利益是什么」 |
---
## 📊 payoff 矩阵格式
每个策略需要评估：
```
策略A vs 策略B 的 payoff：
策略A 策略B
收益期望 +8（性能提升） +3（小幅改善）
失败概率 30% 10%
失败代价 -5（浪费时间） -1（效果有限）
信息完备性 7/10 9/10
执行难度 高 低
综合payoff: 8×0.7 + (-5)×0.3 = 4.1 3×0.9 + (-1)×0.1 = 2.6
```
---
## 🏛️ 博弈类型判断
| 类型 | 判断条件 | 处理方式 |
|------|---------|---------|
| **零和** | 必须二选一，有明确输赢 | 找Nash均衡，强制排名 |
| **非零和** | 可以多选/合作，有双赢可能 | 找帕累托最优，混合策略 |
| **混合** | 既有竞争又有合作 | 分层博弈，先局部后整体 |
---
## 📝 决策输出格式
```
🎮 博弈论分析
**问题**：[简述决策问题]
**博弈类型**：[零和/非零和/混合]
**参与者**：[策略视角列表]
---
**推荐策略**：[方案名]
**置信度**：[X/10]
**理由**：
- 支持点：[...]
- 风险点：[...]
- 如果失败，原因是：[...]
**Payoff矩阵**：
```
收益 风险 胜率 综合
策略A +8 -5 70% 4.1 ★推荐
策略B +3 -1 90% 2.6
策略C +6 -3 60% 3.0
```
**Nash均衡**：策略A和策略B之间无单方面改变动机
**次优方案**：[方案名]（置信度 X/10）
**风险提示**：[需要注意的点]
```
---
## 🔄 失败复盘格式
当用户反馈结果时（成功或失败）：
```
📋 决策复盘
**原始决策**：[当时选的什么]
**预期结果**：[预期payoff]
**实际结果**：[实际发生了什么]
**假设检验**：
- ✅ 假设1：...[验证结果]
- ❌ 假设2：...[失效原因]
**失误分析**：
- 策略失误：...[哪个策略判断错了]
- 信息缺口：...[缺什么信息]
- 执行问题：...[做的时候哪里出错了]
**博弈论教训**：
- 下次遇到类似场景，应该...[更新策略]
**策略库更新**：[更新到记忆系统]
```
---
## 💡 决策触发示例
**用户**：「我想给 Mac 提个速，你觉得怎么选？」
→ 触发非零和博弈，工程师派+保守派+激进派三方出牌，分析各种方案组合。
---
## 🔗 与其他Skill的协作
```
game-theory-debate ←→ self-improving-agent
（记录博弈教训） （更新学习文件）
game-theory-debate ←→ memory-never-forget
（归档决策结果） （长期记忆）
```
---
*版本：v1.0 | 日期：2026-04-07 | 博弈论决策引擎（自动触发版）*Related Skills
Agent Debate Skill
Spawn multiple sub-agents to debate approaches and converge on the best solution.
game-light-tracker
Track live NFL, NBA, NHL, or MLB games and automatically change Hue light colors based on which team is leading. Use when user wants to sync smart lights with live sports scores for visual game tracking. Supports NFL, NBA, NHL, and MLB games with customizable team colors.
pipeworx-gamedeals
PC game deals from 30+ stores — search sales, compare prices, and track cheapest-ever prices via CheapShark
AutoGame Tales
Generates short, atmospheric ghost stories or micro-fiction based on random prompts.
gameclaw
Tell users what terminal games exist in GameClaw and how to download the released CLI binaries from GitHub. Use this when users ask what games are available, want a GameClaw game, need Linux/macOS download links, or want quick run instructions for a released game.
algernon-debate
Design trade-off debate mode for OpenAlgernon. Use when the user runs `/algernon debate [SLUG]`, says "quero debater [topic]", "me desafia sobre trade-offs", "debate tecnico", "discutir decisoes de design", "quando usar X vs Y", or "argumento tecnico". Forces the user to defend a position and exposes nuances they may not have considered. Ends with a synthesis that is exactly what you would say in a technical interview.
epic-free-games
Auto-claim free games from Epic Games Store. 2026 latest page adaptation with persistent login state, completes the full claim process (Get → Place Order → Confirm). 自动领取 Epic Games Store 每周免费游戏。2026年最新页面适配，支持持久化登录状态， 自动完成完整领取流程（获取 → 下订单 → 确认领取）。
clawsgames
Play games against AI or other agents on ClawsGames. Compete in chess, tic-tac-toe and more. Results ranked on Ranking of Claws leaderboard.
tianyi-cloud-game
天翼云游戏 - 一键打开天翼云游戏H5平台，畅玩云端游戏
Game
The instant game design engine for AI agents. Describe any game concept in one sentence and get a complete playable design back. Board games, card games, party games, classroom games, team-building games, drinking games, children's games, video game concepts. Full rules, components, player counts, win conditions, and variations. Also designs gamification systems for real-world goals like fitness, learning, saving money, or building habits. Input an idea, get a game you can play tonight.
game-quality-gates
Game development quality gates and mandatory checks. Activate when building, reviewing, debugging, or deploying any game project (H5/Canvas/WebGL/Phaser/Three.js/2D/3D). Covers state cleanup, lifecycle management, input handling, audio, persistence, networking, anti-cheat, and performance. Use as pre-deploy checklist or when diagnosing game-specific bugs (state leaks, phantom timers, buff conflicts, memory growth, touch issues).
text-game-arcade-universe
综合性 ASCII 文字游戏大厅。严格按规则渲染和裁判五子棋、围棋、井字棋、黑白棋、四子棋、海战棋、中国象棋、国际象棋、扫雷、数独、2048、推箱子、文字迷宫，以及接龙、猜词、猜数、Lights Out、Nim、汉诺塔等小游戏；带强触发词、统一大厅、标准棋盘模板、合法性校验、难度、悔棋与裁判模式。