Get Prompt

Retrieves a prompt from an MCP server with optional parameters and returns the rendered prompt content. This node enables your automation to fetch dynamic prompt templates from MCP servers.

Common Properties

• Name - The custom name of the node.
• Color - The custom color of the node.
• Delay Before (sec) - Waits in seconds before executing the node.
• Delay After (sec) - Waits in seconds after executing node.
• Continue On Error - Automation will continue regardless of any error. The default value is false.

Inputs

• Client Id - The MCP client connection identifier from the Connect node.
• Prompt Name - The name of the prompt to retrieve from the server.
• Parameters - JSON object containing parameters to pass to the prompt template (optional). Parameters are substituted into template variables.

Options

• Timeout - Timeout in seconds for the operation (default: 30).

Outputs

• Prompt Content - The rendered prompt content as a string, with parameters substituted into the template.

How It Works

The Get Prompt node retrieves prompts from MCP servers:

1. Client Lookup - Retrieves the MCP client using the provided Client ID
2. Parameter Processing - Converts parameters to map[string]string format
3. Request Creation - Creates an MCP GetPromptRequest with name and parameters
4. Prompt Retrieval - Sends request to server and receives prompt response
5. Content Extraction - Extracts text content from the prompt messages
6. Output - Returns the rendered prompt content

Requirements

• Valid Client ID from a Connect node
• Non-empty Prompt Name
• Active connection to the MCP server
• Optional parameters matching the prompt's expected variables

Error Handling

The node returns specific errors:

• ErrInvalidArg - Client ID or Prompt Name is empty
• ErrMCPToolCall - Failed to retrieve the prompt from the server

Usage Notes

Parameters Format

Parameters can be provided in multiple formats:

• JSON Object: {"name": "Alice", "company": "Acme Corp"}
• JSON String: '{"name": "Alice"}'
• Message Variable: msg.prompt_params

Parameter Conversion

• All parameter values are converted to strings
• Non-string values are JSON-serialized
• Empty parameters are treated as empty object {}

Prompt Content

• The output is the rendered prompt text
• Template variables are replaced with parameter values
• If multiple messages exist, the first message's text is returned

Examples

Example 1: Simple Greeting Prompt

Inputs:

• Client Id: prompt-server
• Prompt Name: greeting
• Parameters:

{
  "name": "John",
  "company": "Tech Corp"
}

Prompt Content Output:

Welcome John to Tech Corp! We're excited to have you join our team.

---

Example 2: Code Review Prompt

Inputs:

• Client Id: ai-prompts
• Prompt Name: code_review
• Parameters:

{
  "language": "Python",
  "focus": "security",
  "severity": "critical"
}

Prompt Content Output:

Review the following Python code for:
- Best practices
- Security issues
- Performance optimization
- Code style consistency

Focus area: security
Severity level: critical

---

Example 3: Email Template

Inputs:

• Client Id: template-server
• Prompt Name: email_template
• Parameters:

{
  "recipient_name": "Alice",
  "subject": "Project Update",
  "sender_name": "Bob",
  "sender_title": "Project Manager"
}

Prompt Content Output:

Subject: Project Update

Dear Alice,

[Body content would be here]

Best regards,
Bob
Project Manager

---

Example 4: Report Generation Prompt

Inputs:

• Client Id: report-server
• Prompt Name: sales_report
• Parameters:

{
  "period": "Q1 2024",
  "region": "North America",
  "metric": "revenue"
}

Prompt Content Output:

Generate a sales report for Q1 2024 focusing on North America.
Primary metric: revenue
Include trend analysis and comparison with previous period.

---

Example 5: No Parameters

Inputs:

• Client Id: prompt-server
• Prompt Name: daily_standup
• Parameters: (empty)

Prompt Content Output:

Daily Standup Questions:
1. What did you accomplish yesterday?
2. What are you working on today?
3. Are there any blockers?

Best Practices

1. Prompt Discovery:

   • Use List Prompts node first to discover available prompts
   • Verify prompt name spelling matches exactly
   • Check prompt descriptions for parameter requirements

2. Parameter Management:

   • Provide all required parameters
   • Use consistent parameter naming
   • Document parameter purposes
   • Handle optional parameters appropriately

3. Error Handling:

   • Check if prompt retrieval succeeded
   • Validate prompt content is not empty
   • Handle missing prompts gracefully

4. Response Processing:

   • Parse the prompt content as needed
   • Use the content in subsequent AI operations
   • Store prompts for reuse if appropriate

5. Timeout Configuration:

   • Use default timeout for simple prompts
   • Increase timeout for complex prompt generation
   • Handle timeout errors

Common Errors and Solutions

Error: "Client Id cannot be empty"

Cause: Client ID input is missing or empty.

Solution:

• Connect Client Id from a Connect node
• Verify Connect node executed successfully
• Check Client Id is being passed correctly

Error: "Prompt name cannot be empty"

Cause: Prompt Name input is missing or empty.

Solution:

• Provide a valid prompt name
• Use List Prompts to discover available prompts
• Check spelling matches exactly

Error: "MCP client not found"

Cause: No client exists with the provided Client ID.

Solution:

• Ensure Connect node executed before Get Prompt
• Verify Client ID is correct
• Check connection didn't fail or close

Error: "Failed to get prompt"

Cause: Prompt retrieval failed on the server.

Solution:

• Verify prompt exists on the server
• Check parameters match prompt's expected variables
• Ensure prompt name is correct
• Check server logs for errors

Tips for Effective Use

1. List First - Use List Prompts to discover available prompts
2. Understand Parameters - Know which parameters each prompt expects
3. Handle Errors - Implement error handling for failed retrievals
4. Cache Prompts - Cache static prompts to reduce server calls
5. Validate Output - Check prompt content is not empty
6. Use with AI - Feed prompt content to AI models (Claude, GPT, etc.)
7. Test Parameters - Test with various parameter combinations

Integration with AI Models

Pattern: Using Prompts with Claude

Get Prompt → Claude Generate Text

Flow:

1. Get Prompt retrieves template
2. Prompt content becomes User Prompt or System Prompt
3. Claude generates response based on the prompt

Example:

Get Prompt (code_review) → Claude Generate Text
                            Input: code to review
                            System: [prompt content]

Pattern: Dynamic Prompt Selection

Determine Context → Get Prompt (dynamic name) → AI Model

Flow:

1. Logic determines which prompt to use
2. Get Prompt retrieves appropriate prompt
3. AI model processes with the prompt

Usage Patterns

Pattern 1: Static Prompt Retrieval

Connect → Get Prompt → Use Prompt Content

Pattern 2: Parameterized Prompt

Prepare Parameters → Get Prompt → AI Processing

Pattern 3: Multi-Prompt Workflow

Connect → Get Prompt A → Process
        → Get Prompt B → Process
        → Combine Results

Pattern 4: Conditional Prompt Selection

Determine Scenario → Branch → Get Prompt X
                            → Get Prompt Y

Pattern 5: Prompt with AI

Get Prompt → Claude/GPT → Process Response

Advanced Examples

Dynamic Parameter Building

// In a JavaScript node before Get Prompt
msg.prompt_params = {
  user_name: msg.user.name,
  user_role: msg.user.role,
  context: msg.current_task,
  date: new Date().toISOString().split('T')[0],
  preferences: JSON.stringify(msg.user.preferences)
};

Multi-Step Prompt Composition

Get Prompt (intro) → Store
Get Prompt (body) → Store → Combine → AI Model
Get Prompt (conclusion) → Store

Localized Prompts

// Select prompt based on language
const language = msg.user.language;
const promptName = `greeting_${language}`; // greeting_en, greeting_es, etc.