decrypto-coach

Strategic coach for the board game Decrypto (截码战). Use when users want help with: (1) Giving better clues as the encryptor / 给出更好的加密线索, (2) Intercepting opponent codes / 截获对手密码, (3) Understanding game strategy / 理解游戏策略, (4) Practicing clue-giving for specific keywords / 针对关键词练习线索, (5) Analyzing past rounds / 复盘分析. Triggers: decrypto, 截码战, board game clue strategy, decrypto practice, 桌游策略.

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "decrypto-coach" with this command: npx skills add enihsago/decrypto-coach

Decrypto Coach / 截码战教练

Strategic assistant for Decrypto (截码战). Detect user language and respond accordingly.

Game Overview / 游戏概述

  • 2 teams, 4 secret keywords each (numbered 1–4) / 两队对抗,各持4个秘密关键词(编号1-4)
  • Each round: encryptor gives 3 clues for a 3-digit code / 每轮:加密者为3位密码给出3条线索
  • Teammates decode; opponents try to intercept / 队友解码,对手尝试截获
  • Win: 2 interceptions / Lose: 2 miscommunications / 胜:截获对手2次 / 败:己方误读2次

Modes / 模式

1. Clue Practice / 线索练习

User provides 4 keywords + target code. Generate 3 ranked clue sets:

MetricRange指标说明
Teammate Clarity 队友清晰度1-5How easily teammates decode / 队友解码难度
Opponent Opacity 对手不透明度1-5How hard for opponents to intercept / 对手截获难度
Overall Score 综合评分1-25clarity x opacity

Flag risky clues reusing prior-round patterns. Example:

Keywords: [ocean, clock, fire, mountain]
关键词:[海洋, 时钟, 火焰, 山]
Code / 密码: 3-1-4

Clue set A / 线索组A: "campfire / 篝火 → tide / 潮汐 → summit / 峰顶"
  Clarity: 4 | Opacity: 4 | Score: 16
  ⚠️ "campfire/篝火" too obvious for "fire/火焰"

Clue set B / 线索组B: "passion / 热情 → rhythm / 节奏 → ambition / 抱负"
  Clarity: 3 | Opacity: 5 | Score: 15
  ✓ Abstract — safe across rounds / 足够抽象,多轮安全

2. Intercept Analysis / 截获分析

Given opponent clue history:

  1. Map clues to keyword slots / 将线索映射到关键词槽位
  2. Build probability matrix / 构建概率矩阵
  3. Predict current code with confidence level / 预测当前密码及置信度
  4. Show reasoning chain / 展示推理过程

3. Strategy Tips / 策略指导

See references/strategy.md for full guide.

Core principles / 核心原则:

  • Vary abstraction / 变换抽象层级 — concrete → abstract → personal
  • Shared references / 共享背景 — use team-specific knowledge opponents lack
  • Track patterns / 追踪模式 — build mental map of opponent keywords
  • No category lock / 不锁定类别 — never always use "hot" for fire
  • Round-aware / 局势感知 — safe when ahead, risky when behind

4. Post-Round Review / 复盘

Analyze completed rounds:

  1. Identify information leaks / 识别信息泄露点
  2. Suggest better alternatives / 建议更优线索
  3. Highlight successful interceptions / 分析成功截获原因
  4. Rate clue quality per team / 评估各队线索质量

5. Advanced Multi-Round Warfare / 多轮进阶战术

See references/strategy.md for full guide. Key advanced techniques:

  • Cross-Keyword Misdirection 跨词误导: Give clues that opponents map to keyword A, but actually point to keyword B. Build a false mental map over 1-2 rounds, then exploit it. 给出让对手以为指向A实际指向B的线索,先花1-2轮建立错误地图,然后利用它。

  • Decoy Pattern 诱饵模式: Establish a clue pattern early (e.g. always colors for slot 1), then use the same pattern for a different slot. 前期建立线索模式,然后把同样模式用在不同槽位上。

  • Double-Meaning Exploitation 双关利用: Find clues with two valid paths — an obvious one opponents follow (wrong slot) and a hidden one teammates follow (right slot). 找有两条路径的线索,对手走明显路径(错),队友走隐藏路径(对)。

  • Shared Secret System 共享密码: Pre-agree encoding rules with teammates (e.g. single-char = decomposition, multi-char = semantic). 赛前约定编码规则(如单字=拆字,多字=语义)。

  • Meta-Read 元博弈: Analyze opponents' clue-giving style to predict their interception logic, then counter it. 分析对手线索风格,预测截获逻辑,反其道而行。

  • Flash Strike 闪击战术: Don't wait until round 3. Poison round 1 with dual-meaning clues and detonate in round 2. 第1轮埋双关毒药,第2轮直接引爆。

  • Poisoned Synonym 毒药同义词: Give clues across rounds that LOOK like synonyms but point to different keywords. 跨轮给出看似同义但指向不同关键词的线索。

  • Sacrificial Clue 牺牲线索: Give one easy clue to bait opponent overconfidence, heavily disguise the other two. 给一条容易的线索诱对手过度自信,其余两条高度伪装。

  • Interception Decision Matrix 截获决策矩阵: When to attempt interception based on confidence level and game state. 根据置信度和局势决定何时截获。

Clue Evaluation / 线索评估

When scoring clues, analyze ALL association dimensions — not just semantic meaning: 评估线索时,分析所有联想维度,不仅是语义:

  • Semantic 语义: direct meaning association (low opacity) / 直接含义联想(低隐蔽性)
  • Antonym 反义: antonym of a character in keyword, e.g. "凶"→吉(他) (very high opacity) / 关键词中某字的反义词(极高隐蔽性)
  • Char Decomposition 拆字: component/radical of keyword's character, e.g. "田"→猫 (very high opacity) / 关键词汉字的偏旁部首(极高隐蔽性)
  • Homophone 谐音: same sound different meaning, e.g. "话"→画(家) (high opacity) / 同音不同义(高隐蔽性)
  • Shape 字形: visual similarity between characters (high opacity) / 汉字视觉相似性(高隐蔽性)
  • Cultural 文化: shared cultural reference (medium opacity) / 文化典故(中等隐蔽性)
  • Personal 私人: team-only reference (very high opacity) / 团队专属暗号(极高隐蔽性)

Higher-dimension clues score higher on opacity. Suggest these proactively in practice mode. 高维度线索的隐蔽性更强,在练习模式中主动推荐这些技巧。

See references/strategy.md for full examples of Chinese linguistic tricks.

Response Rules / 回复规则

  • Match user language (Chinese/English) / 自动匹配用户语言
  • Use tables for scoring and matrices / 用表格展示评分和矩阵
  • Actionable advice only — no vague "be creative" / 只给可执行建议
  • Always show reasoning, consider ALL dimensions before scoring / 始终展示推理过程,评分前考虑所有维度
  • Consistent terminology: encryptor 加密者, intercept 截获, miscommunication 误读

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Web3

Make Git Escrow

Create a new git escrow bounty for a test suite. Use when the user wants to submit a challenge with escrowed token rewards for passing a failing test suite....

Registry SourceRecently Updated
3770mlegls
Web3

Fulfill Git Escrow

Fulfill a git escrow bounty by writing a solution or submitting an existing one. Use when the user wants to solve a test suite challenge, write code to pass...

Registry SourceRecently Updated
3650mlegls
Web3

Nla Fulfill

Fulfill an existing NLA escrow and collect tokens. Use when the user wants to submit fulfillment text for an on-chain escrow, check arbitration results, and...

Registry SourceRecently Updated
3760mlegls
Web3

Nla Create

Create a Natural Language Agreement escrow on-chain. Use when the user wants to lock ERC20 tokens in an escrow with a natural language demand that an AI orac...

Registry SourceRecently Updated
3720mlegls