Back to Details

cognitive-pattern

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Cognitive Pattern

认知模式

Define cognitive patterns that guide agent reasoning and decision-making.
定义用于指导Agent推理与决策的认知模式。

When to use

适用场景

When you want to codify reasoning strategies, decision trees, or problem-solving approaches that agents can use and improve over time.
当你想要将Agent可使用并随时间迭代优化的推理策略、决策树或问题解决方法进行编码时。

Steps

步骤

  1. Define pattern — call 
    mcp__claude-flow__daa_cognitive_pattern
    with the pattern structure and triggers
  2. Create workflow — call 
    mcp__claude-flow__daa_workflow_create
    to embed the pattern in an executable workflow
  3. Execute — call 
    mcp__claude-flow__daa_workflow_execute
    to run the cognitive workflow
  4. Evaluate — call 
    mcp__claude-flow__daa_learning_status
    to see how the pattern performs
  5. Share — call 
    mcp__claude-flow__daa_knowledge_share
    to propagate successful patterns
  1. 定义模式 — 调用
    mcp__claude-flow__daa_cognitive_pattern
    ，传入模式结构与触发条件
  2. 创建工作流 — 调用
    mcp__claude-flow__daa_workflow_create
    ，将模式嵌入可执行工作流中
  3. 执行 — 调用
    mcp__claude-flow__daa_workflow_execute
    ，运行认知工作流
  4. 评估 — 调用
    mcp__claude-flow__daa_learning_status
    ，查看模式的表现情况
  5. 共享 — 调用
    mcp__claude-flow__daa_knowledge_share
    ，推广成功的模式

Pattern types

模式类型

  • Decision trees — structured if/then reasoning
  • Chain of thought — step-by-step problem decomposition
  • Analogy mapping — solve new problems by referencing similar solved ones
  • Consensus patterns — multi-perspective evaluation before deciding
  • 决策树 — 结构化的if/then推理方式
  • 思维链 — 分步式问题分解
  • 类比映射 — 通过参考相似的已解决问题来解决新问题
  • 共识模式 — 决策前进行多视角评估