MCP-Server

RifatXia/MCP-Server

3.2

If you are the rightful owner of MCP-Server 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.

The Model Context Protocol (MCP) Server is a versatile server designed to handle various data processing tasks using different handlers and tools. It is built using FastAPI and supports JSON-RPC for communication.

Tools
4
Resources
0
Prompts
0

MCP Server Implementation

Student Information

  • Name: Zia Uddin Chowdhury
  • Student ID: A20615319

MCP Capabilities

Parquet Handler

  • Reads specific columns from .parquet files using PyArrow (pyarrow must be installed)
  • Expects valid file paths and column names
  • Supports error handling for:
    • Non-existent files
    • Missing or incorrect column names

Sort Handler

  • Sorts log file lines by timestamps in the format: YYYY-MM-DD HH:MM:SS
  • Handles edge cases such as:
    • Empty files
    • Invalid timestamp formats
  • Returns either sorted entries or a descriptive error message

Compression Handler

  • Compresses files into .gz format using gzip
  • Outputs compressed files to the same directory as the original
  • Preserves the original files
  • Provides compression statistics:
    • Original size
    • Compressed size
    • Compression ratio
  • Skips compression for empty files and raises a warning/error

Pandas Handler

  • Analyzes .csv files using pandas (headers required)
  • Filters rows based on numeric column thresholds
  • Returns results in a structured format
  • Handles missing values and malformed files gracefully

Environment Setup (Linux)

  1. Install uv:
pip install uv
  1. Create and activate environment using uv:
uv venv mcp-server
source mcp-server/bin/activate
  1. Install dependencies using uv:
# Install dependencies from pyproject.toml
uv pip install --requirement pyproject.toml

Dependencies are specified in pyproject.toml:

[project]
name = "mcp_server"
version = "0.1.0"
dependencies = [
    "fastapi>=0.115.12",
    "numpy>=2.2.4",
    "pandas>=2.2.3",
    "pyarrow>=19.0.1",
    "pydantic>=2.11.3",
    "pytest>=8.3.5",
    "pytest-asyncio==0.26.0",
    "requests>=2.32.3",
    "uvicorn>=0.34.1",
]

python = ">=3.10"

Running the MCP Server

  1. Ensure you're in the virtual environment
  2. Start the FastAPI server using uvicorn:
uvicorn src.server:app --reload --host 0.0.0.0 --port 8000

The server will start on http://localhost:8000 by default.

Running Tests

Run all tests:

python3 -m pytest

Successful Tests

Run specific test files:

python3 -m pytest tests/test_sort_handler.py
python3 -m pytest tests/test_parquet_handler.py
python3 -m pytest tests/test_compression_handler.py
python3 -m pytest tests/test_pandas_handler.py

Project Structure

MCP-Server/
ā”œā”€ā”€ src/
ā”‚   ā”œā”€ā”€ capabilities/
ā”‚   ā”‚   ā”œā”€ā”€ __init__.py
ā”‚   ā”‚   ā”œā”€ā”€ compression_handler.py
ā”‚   ā”‚   ā”œā”€ā”€ pandas_handler.py
ā”‚   ā”‚   ā”œā”€ā”€ parquet_handler.py
ā”‚   ā”‚   ā””ā”€ā”€ sort_handler.py
ā”‚   ā”œā”€ā”€ __init__.py
ā”‚   ā”œā”€ā”€ mcp_handlers.py
|   ā””ā”€ā”€ server.py
ā”œā”€ā”€ tests/
ā”‚   ā”œā”€ā”€ __init__.py
ā”‚   ā”œā”€ā”€ test_compression_handler.py
ā”‚   ā”œā”€ā”€ test_pandas_handler.py
ā”‚   ā”œā”€ā”€ test_parquet_handler.py
ā”‚   ā””ā”€ā”€ test_sort_handler.py
ā”œā”€ā”€ data/
ā”œā”€ā”€ images/
ā”œā”€ā”€ README.md
ā”œā”€ā”€ pyproject.toml
ā””ā”€ā”€ pytest.ini

JSON-RPC Requests and Responses

1. List Resources

List Resources Request/Response

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/listResources",
    "id": 1
}

Response:

{
    "jsonrpc": "2.0",
    "result": [
        {
            "id": "resource1",
            "name": "Weather Data",
            "type": "Parquet",
            "description": "Weather measurements including temperature, humidity, and pressure",
            "path": "data/weather_data.parquet",
            "format": "parquet",
            "columns": ["temperature", "humidity", "pressure", "timestamp"]
        },
        {
            "id": "resource2",
            "name": "System Logs",
            "type": "Log",
            "description": "System event logs with timestamps",
            "path": "data/huge_log.txt",
            "format": "text",
            "schema": "timestamp:string message:string level:string"
        },
        {
            "id": "resource3",
            "name": "Student Records",
            "type": "CSV",
            "description": "Student academic records with marks",
            "path": "data/data.csv",
            "format": "csv",
            "columns": ["id", "name", "subject", "marks"]
        },
        {
            "id": "resource4",
            "name": "Application Logs",
            "type": "Log",
            "description": "Application startup and runtime logs with timestamps and log levels",
            "path": "data/output.log",
            "format": "text",
            "schema": "timestamp:string level:string message:string",
            "sample": "[2024-03-16 00:00:15] INFO: Application startup"
        }
    ],
    "id": 1
}

2. Get Resource

Get Resource Request/Response

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/getResource",
    "params": {
        "id": "resource1"
    },
    "id": 1
}

Response:

{
    "jsonrpc": "2.0",
    "result": {
        "id": "resource1",
        "name": "Weather Data",
        "type": "Parquet",
        "description": "Weather measurements including temperature, humidity, and pressure",
        "path": "data/weather_data.parquet",
        "format": "parquet",
        "columns": ["temperature", "humidity", "pressure", "timestamp"]
    },
    "id": 1
}

3. List Available Tools

List Tools Request/Response

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/listTools",
    "id": 1
}

Response:

{
    "jsonrpc": "2.0",
    "result": [
        {
            "id": "tool1",
            "name": "Parquet Reader",
            "description": "Reads columns from Parquet files",
            "usage": "'tool': 'parquet', 'file': 'filename (optional)', 'column': 'column_name' in params."
        },
        {
            "id": "tool2",
            "name": "Parallel Sorting",
            "description": "Sorts log file entries by timestamp",
            "usage": "'tool': 'sort', 'file': 'log_filename' in params."
        },
        {
            "id": "tool3",
            "name": "Compression Tool",
            "description": "Compresses files using gzip",
            "usage": "'tool': 'compress', 'file': 'filename' in params."
        },
        {
            "id": "tool4",
            "name": "Data Analysis using Pandas",
            "description": "Analyzes CSV files using pandas",
            "usage": "'tool': 'pandas', 'file': 'filename', 'column': 'column_name', 'threshold': value in params."
        }
    ],
    "id": 1
}

4. Read Parquet Data

Read Parquet Data Request/Response

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/callTool",
    "params": {
        "tool": "parquet",
        "column": "temperature"
    },
    "id": 2
}

Response:

{
    "jsonrpc": "2.0",
    "id": 2,
    "result": [
        14.96,
        2.01,
        0.56,
        16.19,
        30.18,
        "..."
    ]
}

5. Sort Log Data

Example 1: Small Log File

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/callTool",
    "params": {
        "tool": "sort",
        "file": "small_log.txt"
    },
    "id": 1
}

Response:

{
    "jsonrpc": "2.0",
    "id": 1,
    "result": [
        "2024-03-15 09:15:22 WARNING High CPU usage detected",
        "2024-03-15 09:30:55 WARNING Network latency increased",
        "2024-03-15 10:30:45 INFO Server started successfully",
        "2024-03-15 11:00:45 INFO Scheduled maintenance started"
    ]
}
Example 2: Large Log File

Sort Large Log Request/Response

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/callTool",
    "params": {
        "tool": "sort",
        "file": "huge_log.txt"
    },
    "id": 1
}

Response:

{
    "jsonrpc": "2.0",
    "id": 1,
    "result": [
        "2024-03-15 09:00:00 INFO System initialization complete",
        "2024-03-15 09:15:22 WARNING High CPU usage detected",
        "2024-03-15 09:30:55 WARNING Network latency increased",
        "2024-03-15 09:45:18 INFO User authentication successful",
        "2024-03-15 10:00:15 INFO Backup process started",
        "2024-03-15 10:15:33 ERROR File system error detected",
        "2024-03-15 10:30:45 INFO Server started successfully",
        "2024-03-15 11:00:45 INFO Scheduled maintenance started",
        "2024-03-15 11:30:00 WARNING Memory usage above 80%",
        "2024-03-15 11:45:30 ERROR Database connection failed"
    ]
}

6. Compress File

Compress File Request/Response

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/callTool",
    "params": {
        "tool": "compress",
        "file": "output.log"
    },
    "id": 1
}

Response:

{
    "jsonrpc": "2.0",
    "id": 1,
    "result": {
        "status": "success",
        "original_file": "data/output.log",
        "compressed_file": "data/output.log.gz",
        "original_size": 603,
        "compressed_size": 337,
        "compression_ratio": "44.11%"
    }
}

7. Process CSV Data

Process CSV Data Request/Response

Request:

{
    "jsonrpc": "2.0",
    "method": "mcp/callTool",
    "params": {
        "tool": "pandas",
        "file": "data.csv",
        "column": "marks",
        "threshold": 95
    },
    "id": 1
}

Response:

{
    "jsonrpc": "2.0",
    "id": 1,
    "result": {
        "status": "success",
        "total_rows": 75,
        "filtered_rows": 4,
        "data": [
            {
                "id": 29,
                "name": "Jordan Thomas",
                "subject": "Science",
                "marks": 100
            },
            {
                "id": 32,
                "name": "Jamie Thomas",
                "subject": "Science",
                "marks": 97
            },
            "..."
        ]
    }
}