Prompt Template

Build prompts from templates with variable substitution using {{variable}} syntax.

Overview

Constructs prompt strings by substituting {{variable}} placeholders in a template with values from a provided variables object or from input item fields. Pure string manipulation — no API calls. Supports strict mode that throws on unresolved variables. Each input item produces one rendered prompt stored in a configurable output field (default: “prompt”) along with a list of any unresolved variable names.

Category: AI
Tool Name: prompt_template
Version: 1

Appearance: Icon: brain | Color: #6366f1

Node Type

Action — processes input items and produces output

Input / Output

Direction	Port(s)
Input	Input
Output	Output, Error

Credentials

This tool does not require any credentials.

Parameters

Parameter	Type	Required	Default	Description
Template	string	Yes	—	Template string with {{variable}} placeholders. Falls back to item.template if empty. Supports expressions.
Variables	string	No	—	JSON object of variable name/value pairs. Falls back to item.variables. Any unresolved variables also fall back to matching item fields. Supports expressions.
Options	collection	No	{}	Optional output and validation settings.
— Response Field Name	string	No	prompt	The output field name where the rendered prompt will be stored.
— Strict Mode	boolean	No	false	If true, throw an error when any {{variable}} placeholder cannot be resolved. If false, unresolved placeholders are left as-is.
Include Input	boolean	No	false	Whether to include the original input item fields in the output alongside the rendered prompt.
Max Concurrency	number	No	50	Maximum number of items to process concurrently. Higher values are safe since this is pure string manipulation.

Output Data

One output item per input item. The rendered text is written to the field named by Response Field Name (prompt by default), alongside unresolvedVars. With Include Input off — the default — those two fields are the entire output item; turn it on to merge the original item fields underneath. Binary data is forwarded either way.

{
  "prompt": "Hi Maria, welcome to Acme CRM!",
  "unresolvedVars": []
}

• unresolvedVars lists the placeholder names still present in the rendered text — the ones that matched neither the Variables object nor a field on the input item. It is an empty array when everything resolved.
• With Strict Mode on, an unresolved placeholder becomes an item error instead, so unresolvedVars on a successful item is always empty.

Reference the result downstream by expression, e.g. {{ $json.prompt }}.

Usage Examples

• Build a system prompt with dynamic context variables
• Render a user message template with item-specific values
• Create batch prompts from a template and variable items
• Compose multi-section prompts with topic and content placeholders
• Validate templates in strict mode to catch missing variables

Example Configuration

Minimal configuration — render a simple greeting using variables passed inline:

{
  "name": "Build Greeting",
  "type": "prompt_template",
  "parameters": {
    "template": "Hi {{firstName}}, welcome to {{serviceName}}!",
    "variables": "{\"firstName\": \"Maria\", \"serviceName\": \"Acme CRM\"}",
    "includeInput": false,
    "maxConcurrency": 50
  }
}

Pass-through with input fields — include the original item fields in the output and resolve variables from the item itself, with no explicit variables object:

{
  "name": "Enrich With Prompt",
  "type": "prompt_template",
  "parameters": {
    "template": "Summarize the following ticket (#{{ticketId}}): {{description}}",
    "variables": "",
    "includeInput": true,
    "maxConcurrency": 50,
    "options": {
      "responseFieldName": "prompt"
    }
  }
}

Strict mode with a custom output field — generate an LLM system prompt, store it in systemPrompt, and fail if any placeholder is unresolved:

{
  "name": "Build System Prompt",
  "type": "prompt_template",
  "parameters": {
    "template": "You are a {{role}} assistant working for {{company}}. Always respond in {{language}}.",
    "variables": "{\"role\": \"support\", \"company\": \"Acme Inc\", \"language\": \"English\"}",
    "includeInput": false,
    "maxConcurrency": 20,
    "options": {
      "responseFieldName": "systemPrompt",
      "strictMode": true
    }
  }
}

Dynamic prompts from upstream data — when input items already carry the variable values as fields, omit variables entirely and each {{field}} placeholder resolves directly from the item:

{
  "name": "Prompt From Item Fields",
  "type": "prompt_template",
  "parameters": {
    "template": "Write a product description for {{productName}} priced at {{price}}.",
    "variables": "",
    "includeInput": true,
    "maxConcurrency": 50,
    "options": {
      "responseFieldName": "productPrompt"
    }
  }
}

High-volume batch processing — for large batches of items, raise maxConcurrency since no external I/O is involved:

{
  "name": "Batch Email Subjects",
  "type": "prompt_template",
  "parameters": {
    "template": "Re: {{subject}} — reply needed by {{dueDate}}",
    "variables": "",
    "includeInput": false,
    "maxConcurrency": 200,
    "options": {
      "responseFieldName": "emailSubject",
      "strictMode": false
    }
  }
}

Strict pipeline with controlled variables — in validated pipelines where missing data must halt execution, combine explicit variables with strictMode: true. Here inputText is not in variables, so it is resolved from the input item field inputText; if no such field exists the node raises an error:

{
  "name": "Validated Prompt Builder",
  "type": "prompt_template",
  "parameters": {
    "template": "Translate the following text to {{targetLanguage}}: {{inputText}}",
    "variables": "{\"targetLanguage\": \"French\"}",
    "includeInput": false,
    "maxConcurrency": 50,
    "options": {
      "responseFieldName": "translationPrompt",
      "strictMode": true
    }
  }
}

Error Handling

Mode	Behavior
stop	Halts workflow on first error
continue	Skips failed items, passes successful ones through
errorPort	Routes failed items to Error output port

Tips

The Prompt Template tool substitutes {{variable}} placeholders in a template string with values from a variables object or input item fields, with no external API calls. Use it to dynamically build LLM prompts or structured text that must vary per item before reaching a downstream API or processing node. Each item outputs a rendered prompt in a configurable field (default: prompt) and a list of any placeholder names that could not be resolved.

Key Rules Summary

Rule	Detail
options fields go inside options	Never place responseFieldName or strictMode at the top level of parameters
variables is a JSON string	Serialize the object: "{\"key\": \"value\"}"
Unresolved placeholders fall back to item fields	Explicit variables → item.variables → item fields
strictMode: false (default) leaves unresolved {{vars}} as-is	Set true to error instead
includeInput: true merges input fields into output	Useful when downstream nodes need both the prompt and original data

Placeholder syntax

• A placeholder is matched as {{name}} where the name is letters, digits and underscores only. Write {{topic}}, not {{ topic }} — a placeholder with spaces or dots inside the braces is never substituted, and never appears in unresolvedVars either.
• Values from the Variables object win over fields of the same name on the input item, so an explicit variable always overrides upstream data.