<div align="center">

🏠 Home › Workflows › 🚦 Routing

← 01 Prompt Chaining ━━━━━━━━━━━●━━━━━━━━━━━━━━━━━━━━━━━━ 03 Parallelization →

</div>

---

🚦 Routing

TL;DR: Classify an input and direct it to a specialized handler. Like a train switch — one input takes ONE track.

---

Diagram

%%{init: {'theme': 'base', 'themeVariables': {'lineColor': '#64748b'}}}%%
flowchart TB
    classDef user fill:#6366f1,stroke:#4f46e5,stroke-width:2px,color:#ffffff
    classDef main fill:#8b5cf6,stroke:#7c3aed,stroke-width:2px,color:#ffffff
    classDef subagent fill:#ec4899,stroke:#db2777,stroke-width:2px,color:#ffffff
    classDef idle fill:#94a3b8,stroke:#64748b,stroke-width:2px,color:#ffffff

    INPUT["🙋‍♀️📥 User Request"]:::user --> ROUTER{"🐔🚦 Classify & Route"}:::main

    ROUTER -.->|"Type A"| HA["🐦💤 Handler A"]:::idle
    ROUTER -->|"🐔🪺 Type B"| HB["🐦⚡ Handler B"]:::subagent
    ROUTER -.->|"Type C"| HC["🐦💤 Handler C"]:::idle
    ROUTER -.->|"Unknown"| DEFAULT["🐔💤 Default"]:::idle

    HB -->|"🐦📤"| FINAL["💁‍♀️📤 User Receives"]:::user

---

Key Insight

┌─────────────────────────────────────────────────────────────────────────────┐
│  🚦 ROUTING: Choose ONE branch                                              │
├─────────────────────────────────────────────────────────────────────────────┤
│                                                                             │
│  Logic: if/else, switch/case                                                │
│  Question: "Where should I send this?"                                      │
│  Result: Single output from chosen handler                                  │
│                                                                             │
│  Analogy: Train switch → One train takes ONE track                          │
│                                                                             │
└─────────────────────────────────────────────────────────────────────────────┘

---

Characteristics

---

When to Use

Routing works well for complex tasks where there are distinct categories that are better handled separately, and where classification can be handled accurately.

---

When NOT to Use

• All inputs require same processing
• Classification is unreliable
• Categories overlap significantly

---

<div align="center">

← 01 Prompt Chaining ━━━━━━━━━━━●━━━━━━━━━━━━━━━━━━━━━━━━ 03 Parallelization →

</div>