eagv1-session-05-ajithvcoder

ajithvcoder/eagv1-session-05-ajithvcoder

3.1

If you are the rightful owner of eagv1-session-05-ajithvcoder and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

This document provides a comprehensive overview of a Model Context Protocol (MCP) server designed to perform a series of tasks including reasoning, ASCII conversion, and email functionality.

EAG-Session-05 Assignment

Contents

Requirements

In short write a mcp server that can

  1. Show reasoning with steps its going to do

  2. convert word India to ascii characters, calculate a squared sum of each ascii values

  3. Verify the result

  4. open paint, create a rectange

  5. Add the answer inside the rectangle

  6. Send the answer as email

From EAG course

  • Redo your last assignment (4th one or 3rd one, paint one) but this time:

  • Use ChatGPT or Cursor/Claude or something that qualifies your prompt with the rules mentioned in this Download thisprompt.

  • Show:

    • New final prompt qualified by ChatGPT or Cursor/Clause
    • Based on this new prompt re-do your assignment, share the README.md link for your code
    • Share the YouTube video showing your plugin's use

  • Your assignment cannot be:

    • any summarizer
    • any stock, crypto analyzer, or (can be linking their price changes in the last 1 month based on news)
    • any other simple tool
    • it can be mathematically based query as in the last session, with no problem.. but must have multiple steps involved.

Development Method

Overview

Write a mcp server that can

  1. Show reasoning with steps its going to do
  2. convert word India to ascii characters, calculate a squared sum of each ascii values
  3. Verify the result
  4. open paint, create a rectange
  5. Add the answer inside the rectangle
  6. Send the answer as email

Method

  • Write MCP server functionalities in paint_mcp_server.py to draw rectangle, write text. Make sure you adjust the coordinates in accordance with your system. As they are hardcoded. Write functionality to send email.
  • Write logging functionality instead of print
  • Test it with mcp dev paint_mcp_server.py
  • Now go to talk2mcp.py and adjust the prompts and other functionalites to satisfy the requirement
Usage

Installations

  • pip install -r requirements.txt

Execution

  • python talk2mcp.py

Testing

  • mcp dev paint_mcp_server.py

Learnings

  • Learnt how to create mcp server functions, show reasoning, verification
  • Learnt how to prompt

Results

Prompt validation from chatgpt

{
  "explicit_reasoning": true,
  "structured_output": true,
  "tool_separation": true,
  "conversation_loop": true,
  "instructional_framing": true,
  "internal_self_checks": true,
  "reasoning_type_awareness": true,
  "fallbacks": true,
  "overall_clarity": "Excellent structure: explicit reasoning, tool separation, JSON-enforced output, and verification built-in. Covers reasoning type tagging and fallbacks. Only minor improvement would be clarifying how to handle ambiguous user intent or malformed inputs."
}

Paint App

Paint

Email functionality

Email

Logs