Jakeyshakey Umami MCP Server

🚀 Umami Analytics Service Prompt System Configuration Guide

Umami is a powerful analytics service framework that supports flexible expansion through custom prompts. This guide details how to implement and manage these prompts in src/analytics_service/server.py.

🚀 Quick Start

This guide will lead you through the process of implementing and managing custom prompts in the Umami analytics service. By following the steps below, you can expand the functionality of the service to meet diverse business needs.

✨ Features

• Flexible Customization: Define specific task operation processes through custom prompts.
• Rich Parameter Support: Prompts can accept various input variables.
• Well - structured Message Roles: Use "system", "user", and "assistant" to organize the conversation flow.
• Multiple Content Types: Support "text" and "resource" content types.

📦 Installation

This guide does not cover installation steps. Please refer to the official Umami documentation for installation instructions.

💻 Usage Examples

Basic Usage

Registering Prompts

Register new prompts in the list_prompts() function:

@app.list_prompts()
async def list_prompts():
    return [
        # Other prompts...
        {
            "name": "Your Prompt Name",
            "description": "Your Prompt Description",
            "arguments": [
                {
                    "name": "Parameter Name 1",
                    "description": "Parameter Description",
                    "required": True/False
                },
                {
                    "name": "Parameter Name 2",
                    "description": "Parameter Description",
                    "required": True/False
                }
            ]
        }
    ]

Implementing Prompt Logic

Handle specific prompt requests in the get_prompt() function:

@app.get_prompt()
async def get_prompt(name: str, arguments: Any):
    # Other prompt handling...
    if name == "Your Prompt Name":
        return {
            "messages": [
                {
                    "role": "user",
                    "content": {
                        "type": "text",
                        "text": f"Your prompt template contains {arguments['Parameter Name']}"
                    }
                }
            ]
        }

Advanced Usage

Adding a New Prompt

Add a new prompt in list_prompts():

{
    "name": "Data Analysis Report",
    "description": "Generate a data summary report.",
    "arguments": [
        {
            "name": "startTime",
            "description": "Start time.",
            "required": True
        },
        {
            "name": "endTime",
            "description": "End time.",
            "required": True
        }
    ]
}

Implementing Prompt Handling Logic

Implement the logic in get_prompt():

if name == "Data Analysis Report":
    return {
        "messages": [
            {
                "role": "system",
                "content": {
                    "type": "text",
                    "text": f"Analyze data from {arguments['startTime']} to {arguments['endTime']}."
                }
            },
            {
                "role": "user",
                "content": {
                    "type": "resource",
                    "resource": {
                        "uri": "data://reports", 
                        "text": "Please provide a data summary.",
                        "mimeType": "application/json"
                    }
                }
            }
        ]
    }

📚 Documentation

Core Concepts of the Prompt System

• Prompt: Defines the operation process for a specific task.
• Parameter: Input variables that a prompt can accept.
• Message Role: Includes "system", "user", and "assistant" for organizing the conversation flow.
• Content Type: Supports "text" and "resource" types.

Message Structure Specification

Role Definition

• role: "system": Used to set the context or provide high - level instructions.
• role: "user": Simulates user input or questions.
• role: "assistant": Represents Claude's response or instructions.

Content Type Specification

• type: "text": For plain text content.
• type: "resource": Contains external resources and requires uri, text, and mimeType attributes:

{
    "role": "user",
    "content": {
        "type": "resource",
        "resource": {
            "uri": "logs://recent",    // Resource identifier
            "text": "[2024-03-14 15:32:11] Error: Connection timed out",
            "mimeType": "text/plain"   // MIME type
        }
    }
}

Best Practices

• Be Specific: Ensure that prompts are targeted at specific tasks.
• Clear Parameter Validation: Define clear input requirements.
• Use Reserved Terms: For example, Prompts is recommended to be translated as "Prompts" or "Prompt System".
• Keep Code Examples Intact: Ensure that variable names and syntax are correct.

Advanced Features

• Error Handling: Add exception catching in get_prompt() to ensure clear feedback when a prompt fails.
• Logging: Record all prompt requests and responses for easy debugging and analysis.

By following this guide, you can flexibly expand the functionality of the Umami analytics service to meet different business needs.