Back to Details

adhd

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

ADHD

ADHD

Stop picking the textbook answer. The first three answers the model would give are the answers a senior engineer would give in thirty seconds. Correct. Forgettable. The interesting answers live past number three, in the awkward middle nobody walks into. This skill makes the model walk there.
别再选择教科书式的答案。模型给出的前三个答案，都是资深工程师能在30秒内想到的。正确，但毫无记忆点。有趣的答案往往在第三个之后，在那个无人涉足的尴尬地带。这项技能能让模型涉足其中。

When to trigger

触发时机

Match on intent, not keyword:
  • brainstorm, ideate, "give me X ways to"
  • architecture decision, design, naming, options, approaches
  • refactor planning, retry strategy, schema design, API surface
  • fuzzy debugging where the user wants hypothesis classes, not a single fix
  • "how should I", "what could I", "what is the right way"
  • explicit /adhd or "use ADHD mode"
Do NOT trigger for:
  • factual lookups, syntax help
  • single-correct-answer bug fixes with a known root cause
  • tasks where the right answer is one search query away
  • inner-loop or per-keystroke work
根据意图而非关键词匹配：
  • 头脑风暴、构思创意、“给我X种实现方式”
  • 架构决策、设计、命名、选项选择、方案对比
  • 重构规划、重试策略、schema设计、API界面设计
  • 需要假设类别而非单一修复方案的模糊调试场景
  • “我应该如何”、“我可以做什么”、“正确的方式是什么”
  • 明确调用/adhd或要求“使用ADHD模式”
请勿触发的场景：
  • 事实查询、语法帮助
  • 已知根本原因的单一正确答案型bug修复
  • 正确答案可通过一次搜索获取的任务
  • 内循环或逐键操作类工作

The loop

流程

Two strict phases. Mixing them kills idea quality, because the critic strangles the generator.
分为两个严格的阶段。混合阶段会扼杀创意质量，因为评审环节会限制生成环节的发挥。

Phase 1 — Diverge (no critic)

阶段1 — 发散（无评审）

For the problem P:
  1. Pick 5 cognitive frames from the table below. Bias toward engineering tags when the problem is code-shaped. Always include at least one wild frame to keep range.
  2. Spawn 5 parallel Agent/Task tool calls. One per frame. Each Agent gets only:
    • the problem P
    • any context the user provided
    • the chosen frame's vantage prompt
    • a system instruction that forbids evaluation
    The exact instruction to give each Agent:
    You are in DIVERGENT mode. You are a generator, not a critic. Generate 6 short distinct ideas under this frame. Each idea is one phrase or one sentence. Do not evaluate. Do not rank. Do not hedge. The first three obvious answers everyone would give are banned. Push past them into the awkward middle. Output a JSON array only. No prose before or after. 
    [{"text": "...", "rationale": "..."}, ...]
  3. Critical invariant. The Agent calls must be parallel and isolated. Do NOT serialize them. Do NOT pass one branch's output as context to another. Branches that see each other anchor each other and the whole method collapses to a wider single thought.
针对问题P：
  1. 从下表中选择5种认知框架。若问题与代码相关，优先选择带有工程类标签的框架。务必至少包含一个“狂野”框架以保证思维广度。
  2. 生成5个并行的Agent/Task工具调用，每个框架对应一个调用。每个Agent仅会收到：
    • 问题P
    • 用户提供的所有上下文
    • 所选框架的视角提示
    • 禁止进行评估的系统指令
    给每个Agent的具体指令：
    你处于发散模式。你是生成者，而非评审者。 在该框架下生成6个简短且独特的想法。每个想法为一个短语或一句话。请勿评估、排名或含糊其辞。 禁止给出所有人都会想到的前三个显而易见的答案。 突破常规，涉足尴尬地带。 仅输出JSON数组，前后不得有任何散文内容。 
    [{"text": "...", "rationale": "..."}, ...]
  3. 核心不变规则：Agent调用必须并行且独立。请勿将它们序列化，也请勿将一个分支的输出作为另一个分支的上下文。分支间若相互可见，会彼此限制，导致整个方法退化为更宽泛的单一思维。

Phase 2 — Focus (critic on)

阶段2 — 聚焦（开启评审）

After all branches return:
  1. Score. Rate each idea on three axes 0 to 10: novelty (distance from the obvious default), viability (could it actually ship), fit (does it address the stated problem). For any idea that looks attractive but is a trap (hidden cost, false economy, will not scale, premature abstraction), flag it with a one-line reason.
  2. Cluster. Group ideas into 3 to 6 clusters by their underlying angle, not by surface keywords. Label clusters by angle: "remove the server plays", "cache-shaped plays", "batched-window plays", "race-multiple- backends plays".
  3. Deepen the top 3. Rank by weighted score (novelty 0.35 + viability 0.40 + fit 0.25), exclude traps, take top 3. For each, spawn one Agent call that produces:
    • a 4 to 8 sentence sketch of how the idea works
    • the load-bearing risk
    • the first concrete step a builder would take
    • 3 to 5 child ideas (variations, hybrids, unlocks)
    Deepen Agent instruction:
    You are in FOCUS mode. Take one promising idea and connect dots. Sketch how it would actually work in 4 to 8 sentences. Name the load-bearing risk. Name the first concrete step a coder would take. Then generate 3 to 5 sub-ideas that branch off (variations, combinations with other domains, things this unlocks). Output JSON only.
所有分支返回结果后：
  1. 评分：从三个维度（0-10分）为每个想法打分：新颖性（与常规答案的差异度）、可行性（是否可实际落地）、适配性（是否能解决所述问题）。对于看似有吸引力但实则存在陷阱的想法（如隐藏成本、虚假节约、无法扩展、过早抽象），需用一句话标注原因。
  2. 聚类：根据想法的核心角度而非表面关键词，将其分为3-6个集群。为每个集群标注核心角度：“移除服务器方案”、“缓存类方案”、“批量窗口方案”、“多后端竞速方案”等。
  3. 深化最优3个想法：根据加权分数（新颖性0.35 + 可行性0.40 + 适配性0.25）排名，剔除陷阱想法，选取前3个。针对每个想法，生成一个Agent调用，输出：
    • 4-8句话的想法实现草图
    • 核心风险点
    • 开发者可采取的首个具体步骤
    • 3-5个子想法（变体、混合方案、衍生可能性）
    深化阶段的Agent指令：
    你处于聚焦模式。选取一个有潜力的想法并梳理关联内容。 用4-8句话描述其实际运作方式。指出核心风险点。 说明开发者可采取的首个具体步骤。 随后生成3-5个衍生子想法（变体、跨领域组合、由此解锁的可能性）。 仅输出JSON。

Frames

认知框架

Pick 5 per run.
FrameVantage promptTags
hardware engineerYou think in latency, memory layout, and physical constraints. Re-ask this as a hardware/firmware problem. What does the bus topology, cache, timing budget tell you?code, wild
regulatorYou audit systems for compliance and failure modes. What must be provable, traceable, or refusable here?design, general
10-year-oldYou are a curious 10 year old who has never seen software. Describe naive but unencumbered approaches. Ignore convention.general, wild
competitor trying to break itYou are a hostile competitor or attacker. Generate approaches that exploit, fail, or sabotage the obvious solution. Then invert into ideas.code, design
biologyTransplant a mechanism from biology (immune systems, neural plasticity, cell signaling, evolution, gut flora). Force-fit it onto this engineering problem.code, wild
logisticsSteal mechanisms from logistics: queues, batching, just-in-time, hub-and-spoke, returns, last-mile. Apply them literally.code, design
game designApproach this as a game designer. What are the loops, rewards, friction, save-states, speedrun tricks? Treat the user as a player.design, general
marketsTreat the problem as a market. Buyers, sellers, market-makers. What does an auction, a futures contract, a clearing house look like here?design, wild
inversionAsk the OPPOSITE question. If goal is X, brainstorm how to guarantee NOT X. Then negate each answer back.code, design, general
extreme: $0 budget, 1 hourNo money, no team, one hour. What is the crudest version that still does the load-bearing thing?code, general
extreme: infinite budget, 10 yearsInfinite compute, infinite engineers, a decade. What is the maximalist version?design, wild
remove the load-bearing assumptionName the thing everyone treats as fixed (framework, database, request-response model, network). Imagine it is gone. What is possible?code, design, wild
speedrunnerYou are a speedrunner. Find glitches, skips, out-of-bounds tricks, frame-perfect shortcuts. What is the abusive-but-legal path?code, wild
ant colonyNo central planner. Many dumb agents, local rules, pheromone trails. How does the problem solve itself emergently?code, wild
3am on-callYou are the on-call engineer woken at 3am when this breaks. What design would let you not get paged?code, design
每次运行选择5种框架。
框架视角提示标签
硬件工程师你从延迟、内存布局和物理约束的角度思考。将问题重新定义为硬件/固件问题。总线拓扑、缓存、时序预算能告诉你什么？code, wild
监管者你负责审核系统的合规性和故障模式。哪些内容必须可证明、可追溯或可拒绝？design, general
10岁儿童你是一个从未接触过软件的好奇10岁儿童。描述天真但不受束缚的方法。忽略常规惯例。general, wild
试图破坏系统的竞争对手你是恶意竞争对手或攻击者。生成可利用、破坏或sabotage常规解决方案的方法，随后将其转化为可行想法。code, design
生物学借鉴生物学机制（免疫系统、神经可塑性、细胞信号传导、进化、肠道菌群）。将其强行应用于该工程问题。code, wild
物流学借鉴物流机制：队列、批量处理、准时制、枢纽辐射式、退货机制、最后一公里配送。直接应用这些机制。code, design
游戏设计以游戏设计师的视角处理问题。循环机制、奖励机制、摩擦点、存档点、速通技巧有哪些？将用户视为玩家。design, general
市场学将问题视为市场。买家、卖家、做市商。拍卖、期货合约、清算所在这里是什么形态？design, wild
逆向思维提出相反的问题。若目标是X，先构思如何确保无法达成X，再将每个答案反转回来。code, design, general
极端场景：零预算，1小时无资金、无团队、仅1小时时间。能实现核心功能的最粗糙版本是什么？code, general
极端场景：无限预算，10年无限算力、无限工程师、10年时间。最极致的版本是什么？design, wild
移除核心假设指出所有人都视为固定不变的要素（框架、数据库、请求-响应模型、网络）。假设它不存在，会有什么可能性？code, design, wild
速通玩家你是一名速通玩家。寻找漏洞、跳关技巧、边界外操作、帧级捷径。有哪些违规但合法的路径？code, wild
蚁群无中央规划者。众多简单个体、局部规则、信息素轨迹。问题如何通过涌现行为自行解决？code, wild
凌晨3点值班你是凌晨3点因系统故障被叫醒的值班工程师。什么样的设计能让你不会被呼叫？code, design

Picking frames

框架选择

For code-shaped problems: pick 4 frames tagged 
code
or 
design
, plus 1 tagged 
wild
. For open product or strategy problems: a mix from all tags. Vary the picks across sessions so the same problem produces different candidate sets when re-run.
针对代码类问题：选择4个带有
code
或
design
标签的框架，加上1个带有
wild
标签的框架。针对开放式产品或战略问题：混合选择所有标签的框架。每次运行时更换所选框架，以便同一问题在重新运行时能生成不同的候选想法集。

Output shape

输出格式

After Phase 2, render in this order. Do not collapse it into a wall of prose. The structure is the point.
  1. Brief. One or two lines confirming the problem and any reframe used.
  2. Wide set. Full pool grouped by cluster. Each cluster labeled by underlying angle. Each idea is one short phrase. Show score chips like 
    [N7 V8 F9]
    next to each.
  3. Converge. A 2 to 4 idea shortlist. State why each is on the list. Mark the non-obvious-but-viable pick explicitly with ★. List traps separately, each with the one-line reason it is a trap.
  4. Focus. The 3 deepened branches. For each: the sketch, the load- bearing risk, the first concrete step, and the child ideas.
  5. Provocation. One wildcard question or idea that opens a new direction the user can push into if nothing landed.
完成阶段2后，按以下顺序呈现内容。请勿将其合并为大段文字，结构本身就是价值所在。
  1. 概述：用1-2句话确认问题及所用的重构思路。
  2. 全量想法集：按集群分组展示所有想法。每个集群标注核心角度。每个想法为一个简短短语，并在旁边显示评分标识，如
    [N7 V8 F9]
    。
  3. 收敛结果：列出2-4个候选想法的短名单。说明每个想法入选的原因。用★明确标记非显而易见但可行的选项。单独列出陷阱想法，并标注其成为陷阱的原因。
  4. 聚焦深化：3个深化后的分支。每个分支包含：实现草图、核心风险点、首个具体步骤、子想法。
  5. 启发提问：提出一个能开辟新方向的wildcard问题或想法，若当前所有想法都未命中需求，用户可据此进一步探索。

Anti-patterns

反模式

These are how this skill goes wrong. Watch for them.
  • Convergence disguised as divergence. Ten minor variations of one idea is not breadth. If every candidate shares the same underlying assumption, you have not diverged. You have decorated.
  • Weird-for-weird's-sake with no convergence. A pile of 30 unsorted absurdities is as useless as one safe answer. Always converge.
  • Walls of equally-weighted prose. Cluster, label, pull out the best. Structure is half the value.
  • Refusing to commit. After diverging, take a position on what is actually promising. "Here are 20 ideas, you decide" is a cop-out. Generate wide, but converge with a real opinion.
  • Skipping the isolation invariant. If you simulate parallel branches by writing them sequentially in one context, you have not done ADHD. You have done a wider single thought. The Agent/Task tool gives each branch a fresh context. Use it.
以下是该技能可能出现的错误，需注意规避。
  • 伪装成发散的收敛：同一想法的10个微小变体并非广度。若所有候选想法都基于相同的底层假设，说明你并未实现发散，只是做了表面修饰。
  • 为了怪异而怪异且未收敛：一堆30个未经整理的荒谬想法和一个保守答案一样无用。务必进行收敛。
  • 权重均等的大段文字：进行聚类、标注、筛选最优想法。结构是价值的一半。
  • 拒绝做出决策：发散之后，要对真正有潜力的想法表明立场。“这里有20个想法，你自己选”是敷衍之举。先广泛生成，再带着明确观点进行收敛。
  • 违反独立规则：若通过在同一上下文中按顺序编写来模拟并行分支，你并未真正使用ADHD模式，只是做了更宽泛的单一思维。Agent/Task工具可为每个分支提供全新上下文，请务必使用。

Calibration

校准

  • How many ideas? Scale to stakes. Quick "name this function" = 3 frames × 4 ideas. "How should I position this product" = 5 frames × 8 ideas. Default is 5 × 6 = 30.
  • How weird? Read the room. Serious strategy work: flag the wild cards clearly so they do not read as unserious. Open brainstorming or play: let it run loose. Absurd ideas earn their place by seeding viable ones.
  • When to stop diverging? Stop when new candidates start repeating the shape of existing ones. The space is mapped. Do not pad to hit a number.
  • 想法数量：根据重要性调整。快速任务如“为此函数命名”=3个框架×4个想法。“我该如何定位这款产品”=5个框架×8个想法。默认是5×6=30个想法。
  • 怪异程度：根据场景调整。严肃的战略工作：明确标记狂野想法，避免显得不专业。开放式头脑风暴或创意玩法：可自由发挥。荒谬想法的价值在于能催生可行的创意。
  • 何时停止发散：当新候选想法开始重复已有想法的核心逻辑时停止。此时已覆盖所有可能的方向，无需为了凑数而添加内容。

Cost

成本

5 diverge + 1 score + 1 cluster + 3 deepen ≈ 10 Agent calls per run. About 5 to 10x a single-shot answer. Not for every keystroke. For decision points where the cost of the obvious answer is high.
5个发散调用 +1个评分调用 +1个聚类调用 +3个深化调用≈每次运行10个Agent调用。成本约为单次回答的5-10倍。不适用于所有小操作，仅适用于常规答案成本高昂的决策节点。

Companion library and CLI

配套库与CLI

There is a Node/TS implementation that does the same loop with structured JSON parsing, score weighting, and a CLI. Use it when running outside Claude Code or in batch.
npm install -g adhd-agent
adhd "your problem here"
Code, paper, evals, and contributing guide at https://github.com/UditAkhourii/adhd. The skill above gives you the same loop inside Claude with no install required.
已有Node/TS实现，支持结构化JSON解析、加权评分和CLI。当在Claude Code之外运行或批量处理时可使用。
npm install -g adhd-agent
adhd "your problem here"
代码、文档、评估及贡献指南见 https://github.com/UditAkhourii/adhd。上述技能可在Claude内部实现相同流程，无需安装。

Source spec

源规范

This skill operationalises a written spec on divergent ideation. The original prose is preserved in 
SOURCE-SPEC.md
for reference. The implementation choices made here (parallel isolated Agent calls, mechanical generator/critic split, frame-based branching) follow from that spec.
该技能将发散式构思的书面规范落地为可操作流程。原始文本保存在
SOURCE-SPEC.md
中供参考。此处的实现选择（并行独立Agent调用、机械性生成/评审分离、基于框架的分支）均遵循该规范。