Prompt Chain Visualizer

Map multi-step LLM workflows. Name each step, write its prompt, link outputs to inputs, and export the full chain as JSON.

100% browser-based · No data stored

Chain Steps

3 steps·~101 tokens total

User Input

Input from: User Input

Input from step

Prompt — use {User Input} to reference input

~33 tokens

Input from: Research

Input from: Analyze

JSON Export

{
  "chain": [
    {
      "step": 1,
      "name": "Research",
      "inputFrom": "User Input",
      "prompt": "You are a research assistant. Given the topic: {input}, find and summarize the 5 most important facts. Format as a numbered list."
    },
    {
      "step": 2,
      "name": "Analyze",
      "inputFrom": "Research",
      "prompt": "You are an analyst. Given these research findings: {Research}, identify the 3 most significant insights and their business implications."
    },
    {
      "step": 3,
      "name": "Draft Report",
      "inputFrom": "Analyze",
      "prompt": "You are a business writer. Using these insights: {Analyze}, write a concise 300-word executive summary suitable for a C-suite audience."
    }
  ]
}