reasoning-router

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Reasoning Router

推理路由器

Meta-reasoning layer: classify problems, dispatch to the optimal thinking MCP, monitor progress, re-route on stall. Minimize total tokens to a high-confidence conclusion.

Always uses the most advanced model (currently Opus 4.6). Meta-reasoning about how to think is the highest-leverage reasoning task.

元推理层：对问题进行分类，调度至最优思维MCP，监控进度，在思维停滞时重新路由。以最少的总token数得出高置信度结论。

始终使用最先进的模型（当前为Opus 4.6）。关于「如何思考」的元推理是价值最高的推理任务。

Dispatch

调度

`$ARGUMENTS`	Action
Problem or question text	Route — classify and dispatch to optimal thinking method
`compare <problem>`	Compare — run 2-3 strategies on the same problem, pick best
`status`	Status — show current thinking session state
`switch <method>`	Switch — re-route to specified method mid-session
`recap`	Recap — summarize reasoning so far across all methods used
`methods`	Methods — show available methods with when-to-use cheatsheet
Empty	Guide — show routing cheatsheet and token tier table

`$ARGUMENTS`	操作
问题或提问文本	路由 — 分类并调度至最优思维方法
`compare <problem>`	对比 — 针对同一问题运行2-3种策略，选择最优方案
`status`	状态 — 展示当前思维会话状态
`switch <method>`	切换 — 会话中途重新路由至指定方法
`recap`	回顾 — 总结目前使用所有方法的推理过程
`methods`	方法 — 展示可用方法及适用场景速查表
空值	指南 — 展示路由速查表和token层级表

Classification

分类

Run

!uv run python scripts/problem-classifier.py "$ARGUMENTS"

for deterministic pre-scan, then refine with LLM judgment.

运行

!uv run python scripts/problem-classifier.py "$ARGUMENTS"

进行确定性预扫描，再通过LLM判断优化结果。

Axis 1 — Structure (what shape is the reasoning?)

维度1 — 结构（推理的形态？）

Structure	Signals	Primary Method
Decomposable	"break down", multi-hop, sub-questions	`atom-of-thoughts`
Sequential	"step by step", ordered process, debugging	`sequential-thinking`
Branching	"explore options", "what if", alternatives	`cascade-thinking`
Constrained	"given that", requirements, formal specs	`shannon-thinking`
Interconnected	dependencies, circular, complex system	`atom-of-thoughts` (full)
Creative	"brainstorm", "novel", "stuck", "reframe"	`creative-thinking`
Contradictory	paradox, "both true", tension, dilemma	`lotus-wisdom`
Investigative	"why did", root cause, debug, diagnose	`crash`

结构	信号特征	首选方法
可分解	"拆分"、多跳推理、子问题	`atom-of-thoughts`
顺序型	"逐步"、有序流程、调试	`sequential-thinking`
分支型	"探索选项"、"假设"、替代方案	`cascade-thinking`
约束型	"给定条件"、需求、正式规范	`shannon-thinking`
互联型	依赖关系、循环逻辑、复杂系统	`atom-of-thoughts` （完整版）
创意型	"头脑风暴"、"新颖"、"陷入瓶颈"、"重构框架"	`creative-thinking`
矛盾型	悖论、"同时成立"、冲突、两难困境	`lotus-wisdom`
调查型	"为何发生"、根本原因、调试、诊断	`crash`

Axis 2 — Complexity (how deep?)

维度2 — 复杂度（推理深度？）

Complexity	Signals	Depth Control
Simple (1-3 steps)	Short input, single concept, "quick"	Tier 1 ONLY: `aot-light` or `sequential`
Moderate (4-8 steps)	Multiple factors, some uncertainty	Default depth, structure-matched method
Complex (9+ steps)	Cross-domain, high uncertainty	Full depth, extended steps
Wicked (open-ended)	No clear solution, values conflict	Composition patterns allowed

复杂度	信号特征	深度控制
简单（1-3步）	输入简短、单一概念、"快速"	仅使用层级1： `aot-light` 或 `sequential`
中等（4-8步）	多因素、存在不确定性	默认深度，匹配结构的方法
复杂（9步以上）	跨领域、高不确定性	完整深度、扩展步骤
棘手（开放式）	无明确解决方案、价值观冲突	允许使用组合模式

Axis 3 — Domain (what kind?)

维度3 — 领域（问题类型？）

Domain	Method	Why
Engineering/architecture	`shannon-thinking`	Constraints → model → proof
Debugging	`crash`	Tool-aware investigation
Research	`cascade-thinking`	Parallel perspectives
Math/logic	`atom-of-thoughts`	Dependency DAG
Strategy	`think-strategies`	Strategy comparison
Creative/design	`creative-thinking`	Lateral techniques
Philosophy/ethics	`lotus-wisdom`	Non-dual integration
Planning	`sequential-thinking`	Linear progression
Pattern tracking	`structured-thinking`	Persistent memory

领域	方法	原因
工程/架构	`shannon-thinking`	约束→建模→验证
调试	`crash`	工具化调查
研究	`cascade-thinking`	多视角并行分析
数学/逻辑	`atom-of-thoughts`	依赖关系DAG
策略	`think-strategies`	策略对比
创意/设计	`creative-thinking`	横向思维技巧
哲学/伦理	`lotus-wisdom`	非二元整合
规划	`sequential-thinking`	线性推进
模式追踪	`structured-thinking`	持久化记忆

Routing Algorithm

路由算法

Optimize for minimum total tokens to reach a high-confidence conclusion.

1. Run problem-classifier.py for deterministic signals
2. Score each axis (structure × complexity × domain)
3. Apply efficiency gate:
   - Simple: ALWAYS Tier 1 (aot-light or sequential-thinking)
   - Moderate: structure-matched method at default depth
   - Complex: full method with extended depth
   - Wicked: composition patterns allowed
4. Resolve conflicts:
   - Structure > domain (HOW > WHAT)
   - Complexity modulates depth AND tier
   - Ties: ALWAYS prefer lower token cost
5. Select primary + fallback
6. Present: "Using [method] — [why]. Fallback: [method]." (1 line)
7. Invoke with minimum viable totalThoughts
8. Monitor for stalls — escalate only on evidence

优化目标：以最少的总token数达成高置信度结论。

1. 运行problem-classifier.py获取确定性信号
2. 对每个维度打分（结构 × 复杂度 × 领域）
3. 应用效率闸门：
   - 简单问题：始终使用层级1（aot-light或sequential-thinking）
   - 中等问题：匹配结构的方法+默认深度
   - 复杂问题：完整方法+扩展深度
   - 棘手问题：允许使用组合模式
4. 解决冲突：
   - 结构 > 领域（方法 > 内容）
   - 复杂度调节深度及层级
   - 平局：始终选择token成本更低的方法
5. 选择主方法+备用方法
6. 告知用户："正在使用[方法] — [原因]。备用方案：[方法]。"（一行内容）
7. 以最小可行totalThoughts调用方法
8. 监控停滞情况 — 仅在有证据时升级方法

Token Cost Tiers

Token成本层级

Tier 1 (lightest): aot-light, sequential-thinking
Tier 2 (moderate): structured-thinking, cascade-thinking, crash
Tier 3 (heavy):    shannon-thinking, atom-of-thoughts (full), think-strategies
Tier 4 (heaviest): creative-thinking, deep-lucid-3d, lotus-wisdom

层级1（最轻量）：aot-light, sequential-thinking
层级2（中等）：structured-thinking, cascade-thinking, crash
层级3（重量级）：shannon-thinking, atom-of-thoughts（完整版）, think-strategies
层级4（最重）：creative-thinking, deep-lucid-3d, lotus-wisdom

Efficiency Rules

效率规则

Never Tier 3-4 for a problem Tier 1-2 can handle
Start with lowest viable
```
totalThoughts
```
; expand only on
```
needsMoreThoughts
```
```
cascade-thinking
```
: start 2 branches max; add only if coverage insufficient
```
creative-thinking
```
: use
```
timeframe: "quick"
```
unless wicked
```
think-strategies
```
: start with
```
linear
```
or
```
chain_of_thought
```
before heavier strategies

层级1-2可处理的问题，绝不使用层级3-4
以最低可行的
```
totalThoughts
```
启动；仅在收到
```
needsMoreThoughts
```
时扩展
```
cascade-thinking
```
：最多从2个分支开始；仅在覆盖不足时添加分支
```
creative-thinking
```
：除非是棘手问题，否则使用
```
timeframe: "quick"
```
```
think-strategies
```
：在使用更重策略前，先从
```
linear
```
或
```
chain_of_thought
```
开始

Stall Detection

停滞检测

Read

references/stall-detection.md

when any stall signal fires.

4 stall signals:

Confidence plateau — no increase >0.05 in 3+ steps
Circular reasoning — same claim restated
Step overrun — exceeds 2× initial estimate
User signal — "try something else" or
```
/think switch
```

Re-routing by block type:

Block	New Method	Why
Stuck in depth	`cascade-thinking`	Broaden
Too broad	`atom-of-thoughts`	Decompose
Too formal	`creative-thinking`	Reframe
Too abstract	`crash`	Ground in evidence
Contradictions	`lotus-wisdom`	Integrate
Wrong method	Reclassify	Start routing over

Hard limits: Max 3 re-routes per session. Max 3× step estimate. No ping-pong (can't return to a method already re-routed from). After 3 re-routes → recovery protocol: present findings, ask user to guide.

当任何停滞信号触发时，阅读

references/stall-detection.md

。

4种停滞信号：

置信度停滞 — 连续3步以上置信度提升≤0.05
循环推理 — 重复陈述相同观点
步骤超时 — 超过初始预估的2倍
用户信号 — "尝试其他方法" 或
```
/think switch
```

按阻塞类型重新路由：

阻塞类型	新方法	原因
陷入深度僵局	`cascade-thinking`	拓宽思路
过于宽泛	`atom-of-thoughts`	分解问题
过于形式化	`creative-thinking`	重构框架
过于抽象	`crash`	基于证据落地
存在矛盾	`lotus-wisdom`	整合观点
方法错误	重新分类	重新执行路由流程

硬限制： 每个会话最多3次重新路由。最多为初始预估步骤的3倍。禁止来回切换（不能回到已重新路由的方法）。3次重新路由后执行恢复协议：展示现有发现，请用户指导方向。

Composition Patterns

组合模式

Read

references/composition-patterns.md

for full protocols.

Pattern	Sequence	When
Sketch→Detail	`aot-light` → `aot` (full)	Quick outline then deep dive
Diverge→Converge	`creative-thinking` → `sequential`	Generate options then evaluate
Explore→Formalize	`cascade-thinking` → `shannon`	Survey landscape then prove
Decompose→Investigate	`aot` → `crash` (per atom)	Break down then debug each
Analyze→Integrate	`cascade-thinking` → `lotus-wisdom`	Surface contradictions then synthesize
Strategize→Plan	`think-strategies` → `sequential`	Step back then plan forward

Context transfer: At every method transition, produce ≤100 word summary of key findings with confidence scores. Carry claims, drop scaffolding.

When NOT to compose: Simple/moderate complexity, first method is converging, time pressure is high.

如需完整协议，请阅读

references/composition-patterns.md

。

模式	顺序	适用场景
草图→细节	`aot-light` → `aot` （完整版）	快速勾勒框架后深入研究
发散→收敛	`creative-thinking` → `sequential`	生成选项后评估
探索→形式化	`cascade-thinking` → `shannon`	调研全局后验证
分解→调查	`aot` → `crash` （针对每个原子单元）	拆分问题后逐个调试
分析→整合	`cascade-thinking` → `lotus-wisdom`	呈现矛盾后综合
制定策略→规划	`think-strategies` → `sequential`	退一步思考后推进计划

上下文传递： 每次方法切换时，生成≤100字的关键发现摘要及置信度评分。保留核心结论，丢弃辅助脚手架。

不适用组合的场景： 简单/中等复杂度问题、第一个方法正在收敛、时间压力大。

Compare Mode

对比模式

When invoked with

compare <problem>

Classify the problem normally
Select 2-3 candidate methods (primary + 1-2 plausible alternatives)
Run each for 3-5 steps on the same problem
Compare: which produced higher-confidence findings with fewer tokens?
Present comparison table and recommend the winner
Continue with the winning method if user agrees

当通过

compare <problem>

调用时：

正常分类问题
选择2-3种候选方法（主方法+1-2种合理替代方案）
针对同一问题每种方法运行3-5步
对比：哪种方法用更少的token产生了更高置信度的结论？
展示对比表格并推荐最优方案
若用户同意，继续使用获胜方法

Methods Quick Reference

方法速查

When invoked with

methods

or for the routing cheatsheet:

Method	MCP Tool	Tier	Sweet Spot
sequential-thinking	`sequentialthinking`	1	Linear, ordered reasoning
aot-light	`aot-light`	1	Quick atomic sketches
structured-thinking	`capture_thought`	2	Cross-session memory
cascade-thinking	`cascade_thinking`	2	Multi-perspective exploration
crash	`crash`	2	Debugging with tool integration
shannon-thinking	`shannonthinking`	3	Formal proof under constraints
atom-of-thoughts	`aot` + `atomcommands`	3	Deep decomposition with DAG
think-strategies	`think-strategies`	3	Named strategy frameworks
creative-thinking	`discover_techniques` → `plan` → `execute`	4	Lateral thinking (3-step required)
deep-lucid-3d	`analyze_problem`	4	Holistic UCPF analysis
lotus-wisdom	`lotuswisdom` (start: `tag: begin` )	4	Paradox and contradiction

Read

references/capability-map.md

for full tool parameters and workflows.

当调用

methods

或查看路由速查表时：

方法	MCP工具	层级	适用场景
sequential-thinking	`sequentialthinking`	1	线性、有序推理
aot-light	`aot-light`	1	快速原子化勾勒
structured-thinking	`capture_thought`	2	跨会话记忆
cascade-thinking	`cascade_thinking`	2	多视角探索
crash	`crash`	2	集成工具的调试
shannon-thinking	`shannonthinking`	3	约束下的正式验证
atom-of-thoughts	`aot` + `atomcommands`	3	带DAG的深度分解
think-strategies	`think-strategies`	3	命名策略框架
creative-thinking	`discover_techniques` → `plan` → `execute`	4	横向思维（需三步流程）
deep-lucid-3d	`analyze_problem`	4	整体UCPF分析
lotus-wisdom	`lotuswisdom` （启动： `tag: begin` ）	4	悖论与矛盾处理

如需完整工具参数及工作流，请阅读

references/capability-map.md

。

Reference File Index

参考文件索引

File	Read When
`references/capability-map.md`	First invocation or unfamiliar with a tool's parameters
`references/routing-matrix.md`	Resolving routing conflicts, ties, or edge cases
`references/composition-patterns.md`	Problem needs a multi-tool sequence
`references/stall-detection.md`	Stall signal fires or session exceeds expected steps

Load ONE reference at a time. Do not preload.

文件	阅读时机
`references/capability-map.md`	首次调用或不熟悉工具参数时
`references/routing-matrix.md`	解决路由冲突、平局或边缘情况时
`references/composition-patterns.md`	问题需要多工具序列时
`references/stall-detection.md`	停滞信号触发或会话超出预期步骤时

每次仅加载一个参考文件，请勿预加载。

Scope Boundaries

范围边界

In scope: Problem classification, thinking MCP dispatch, stall detection, re-routing, composition pattern selection, method comparison.

Out of scope: Actually performing the reasoning (the thinking MCP does that), running tools or writing code (use appropriate skills), code review (review), research (research), debugging implementation (use crash MCP directly).

包含范围： 问题分类、思维MCP调度、停滞检测、重新路由、组合模式选择、方法对比。

排除范围： 实际执行推理（由思维MCP完成）、运行工具或编写代码（使用对应技能）、代码评审（使用评审功能）、研究（使用研究功能）、实现调试（直接使用crash MCP）。

Canonical Vocabulary

标准术语

Term	Meaning
method	A specific thinking MCP tool (e.g., cascade-thinking)
tier	Token cost classification (1=lightest, 4=heaviest)
structure	The shape of reasoning a problem requires
routing	Selecting the best method for a classified problem
stall	When a method stops making progress (confidence plateau, circular reasoning)
re-route	Switching to a different method mid-session
composition	Chaining 2+ methods sequentially for complex problems
efficiency gate	Override that forces Tier 1 methods for simple problems
context transfer	Summarizing findings when switching between methods

术语	含义
method	特定的思维MCP工具（例如cascade-thinking）
tier	Token成本分类（1=最轻量，4=最重）
structure	问题所需的推理形态
routing	为分类后的问题选择最优方法
stall	方法停止产生进展的状态（置信度停滞、循环推理）
re-route	会话中途切换至其他方法
composition	为复杂问题按顺序串联2种及以上方法
efficiency gate	强制简单问题使用层级1方法的规则
context transfer	切换方法时总结发现的操作

Critical Rules

关键规则

Always use the most advanced model —
```
model: opus
```
. Never downgrade.
Minimize total tokens to conclusion — pick the lightest method that handles the problem. Escalate only on evidence of stall.
Present routing decision before invoking — 1 line: method, why, fallback.
Never invoke without classification — even if user names a method, validate fit.
Carry forward on re-route — summarize findings ≤100 words, don't restart.
Simple gets simple — Tier 1 for simple problems, always. No exceptions.
Token cost breaks ties — equal fit → lower tier wins.
Start minimal, expand on evidence — lowest
```
totalThoughts
```
, expand only when the tool requests more or stall detected.
One method at a time — no simultaneous thinking MCPs. Composition is sequential.
Router ≠ thinker — classify, dispatch, monitor. Don't duplicate reasoning.
Respect tool workflows — creative-thinking: discover→plan→execute. lotus-wisdom: start with
```
tag: begin
```
. aot: use
```
atomcommands
```
for control.
Track token spend — note approximate tokens per session to refine routing.

始终使用最先进的模型 —
```
model: opus
```
。绝不降级。
最小化结论所需总token数 — 选择能处理问题的最轻量方法。仅在有停滞证据时升级。
调用前告知路由决策 — 一行内容：方法、原因、备用方案。
未分类绝不调用 — 即使用户指定方法，也要验证适配性。
重新路由时延续上下文 — 总结≤100字的发现，不要从头开始。
简单问题用简单方法 — 简单问题始终使用层级1方法，无例外。
Token成本打破平局 — 适配性相同时，选择更低层级的方法。
从最小配置开始，按需扩展 — 使用最低的
```
totalThoughts
```
，仅在工具请求更多或检测到停滞时扩展。
同一时间仅使用一种方法 — 不同时运行多个思维MCP。组合模式为顺序执行。
路由器≠思考者 — 仅负责分类、调度、监控。不重复执行推理。
遵循工具工作流 — creative-thinking：discover→plan→execute；lotus-wisdom：以
```
tag: begin
```
启动；aot：使用
```
atomcommands
```
进行控制。
追踪Token消耗 — 记录每个会话的大致Token使用量，优化路由策略。