mcp-schedulizer-server-P1-Redes

BrandonReyes0609/mcp-schedulizer-server-P1-Redes

3.1

If you are the rightful owner of mcp-schedulizer-server-P1-Redes and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The MCP-Schedulizer is a local server designed to optimize task scheduling using the Model Context Protocol (MCP) and JSON-RPC 2.0.

🤖 Project 1 Networks (Schedulizer Server)

This document describes the operation of the host agent developed for Project 1 - Use of an existing protocol (CC3067), which uses an LLM (Claude) to interpret natural language instructions and execute tools through a local MCP server compatible with the official protocol.

Prepared by: Brandon Reyes Morales 22992

Available Tools

Local MCP Server (mcp_schedulizer.py)

Theme: Task management and schedule generation (Schedulizer)

Main functions:

  • Add tasks with duration, priority, and deadline
  • List tasks (list_tasks)
  • Generate an optimized schedule (generate_schedule)
  • Save tasks locally in tasks_db.json

Goal: To be the core of the project, simulating an assistant that organizes your week based on your tasks. It can be extended to prioritize courses, work, rest, etc.

Remote MCP Server (Google Cloud Run)

Theme: Complementary educational services and well-being

Implemented functions:

  • suggest_breaks: suggests breaks between tasks in a schedule
  • daily_quote: returns a motivational quote of the day

Goal: Extend the assistant with functions that support student or worker well-being, reinforcing productivity with motivation and healthy pauses. Serves as an example of a cloud-hosted server.

💬 Conversation Example

Online server

Examples for the remote server on Google Cloud (FastAPI):

Can you suggest breaks if I study for 4 hours?

Give me a motivational quote

Add a task to study networks...

Write to file notes.txt

Suggest_breaks — Pomodoro-style suggestions

If I plan to study 3 hours, how many breaks should I take?

I’ll work for 5 hours straight, can you suggest pauses?

Can you give me a Pomodoro-style routine for 2.5 hours of study?

Give me a list of breaks if I plan a 4-hour study session

I want to organize my 6-hour workday with breaks every 45 minutes

Suggest work and break blocks for a 3-hour session

daily_quote — Motivational quotes

Give me a motivational quote to start my day

Do you have a positive quote to inspire me today?

I want an inspiring quote, please

Motivate me with a quote for studying

Can you give me a quote that helps me focus?

How to verify that the remote server is being used?

You should see something like this in the console:

🔗 Claude API used to interpret user instructions...
🔧 Executing 'suggest_breaks' with: {'hours': 4}
🌐 Using remote server on Google Cloud (endpoint /suggest_breaks)
📬 Response: { ... }

or

🔧 Executing 'daily_quote' with: {}
🌐 Using remote server on Google Cloud (endpoint /daily_quote)

Local server

Examples of input for tasks (add_task):

Add a task to study for the networks final, 90 minutes, deadline 2025-09-30

Add a task to study for the networks exam, 90 minutes, deadline 2025-09-30, high priority, category university

Create a new task called “Read chapter 5 of artificial intelligence,” duration 60 minutes, deadline 2025-10-01, medium priority, category study

Add a task to review algorithms, duration 45 minutes, deadline 2025-09-25, low priority, category university

Add a task to write the networks lab report, 120 minutes, deadline 2025-10-02, category university, high priority

Register a task to review cybersecurity for 30 minutes before September 28, category work, medium priority

Examples to list tasks (list_tasks):

Show me all my tasks

What tasks do I have pending?

I want to see my task list

Examples for the Filesystem MCP server (STDIO)

write_file — Write or create files

Create a file called resumen.txt with the content: OSI model layers and their functions

Write in a file called tareas.txt the following: study networks, program simulator

Save in notes/monday.txt: review Claude and remote server tasks

Create a folder 'reports' and inside save a file called info.txt with the content: MCP working correctly

Write in workspace/week1.txt the following: Monday - study, Tuesday - rest, Wednesday - lab

read_file — Read existing files

Read the file resumen.txt

Show me the content of tareas.txt

Open and read the file called info.txt in the reports folder

I want to see what I wrote in week1.txt

Can you read the file notes/monday.txt?

How to know if the Filesystem server is being used?

When you use one of these commands, your console should display:

🔧 Executing 'read_file' with: {'path': 'workspace/resumen.txt'}
📁 Using filesystem STDIO server (Filesystem MCP)
📬 Response:
{
  "output": {
    "content": "OSI model layers and their functions"
  }
}

Folder and file structure

mcp-schedulizer-server-P1-Redes/
├── docs/                            # General and technical documentation
│   ├── screenshots/                 # Screenshots used in reports
│   │   ├── README_LLM_ES.md        # Write-up about LLM integration
│   │   ├── README_LLM_ES copy.md   # Duplicate/backup version
│   │   └── README.md               # General screenshot instructions
│   ├── captura wireshark 1.png     # Chatbot-server communication capture
│   ├── captura_wireshark_mcp_local.png
│   ├── captura_wireshark_server_remoto.png
│   ├── INFORME.docx                # Formal project report (editable)
│   ├── INFORME.pdf                 # Report in PDF format
│   └── prueba2.pcap                # Network capture file (Wireshark)
│
├── src/                            # Main MCP server source code
│   ├── tools/                      # Tool definitions
│   │   ├── add_task.tool.json
│   │   ├── definition.json
│   │   ├── generate_schedule.tool.json
│   │   ├── get_calendar.tool.json
│   │   ├── list_tasks.tool.json
│   │   └── remove_task.tool.json
│   ├── mcp_schedulizer.py          # Local MCP server that organizes tasks
│   ├── cliente_chatbot_fs_rr.py    # Client interacting with Claude and servers (FS + Remote)
│   └── mcp_log.txt                 # Execution log
│
├── workspace/                      # Local workspace (simulated user storage)
│   ├── university/                 # Example subfolder
│   └── agenda.txt                  # File generated with schedule
│
├── venv_mcp/                       # Python virtual environment
│   └── ... (internal files) ...
│
├── .env                            # Local config environment variables
├── .gitignore                      # Git ignored files
├── env.example                     # `.env` template
├── mcp_log.txt                     # Duplicate log outside `src`
├── README_LLM_EN.md                # English LLM integration guide
├── README_mcp_oficial_servers.md   # Instructions to run official MCP servers (FS, Git, etc.)
├── README.md                       # Main project instructions
├── README(1).md                    # Duplicate/backup README
├── requirements.txt                # Project dependencies

✅ Prerequisites

  • Python 3.12+
  • Libraries: anthropic, requests, flask
  • Valid Anthropic API Key
  • MCP server running at http://localhost:5000
pip install flask
pip install requests
pip install rich
pip install anthropic

pip install anthropic python-dotenv requests mcp
pip install flask anthropic python-dotenv requests mcp

📌 Notes

  • This agent does not use menus, it acts as a natural conversational LLM.
  • Claude must respond with exactly valid JSON, so a cleaning system was implemented for json ... responses from the model.