📊 MCP Log Analyzer

An intelligent log file analyzer powered by Python and integrated with Cursor IDE through the Model Context Protocol (MCP).

🎯 What It Does

Analyzes log files instantly using AI-powered tools. Instead of manually searching through thousands of log lines, ask questions in natural language and get instant insights about errors, patterns, and anomalies.

Before vs After

Before (❌):

• Manually open and search log files

• Hours of tedious analysis

• Error-prone manual work

After (✅):

• Ask Claude in natural language

• Get instant analysis in seconds

• Automatic pattern detection

• AI-powered insights

Related MCP server: Local Logs MCP Server

🛠️ Features

6 Analysis Tools

1. read_logs - Fetch and display log content

2. count_log_types - Count logs by severity (INFO, ERROR, WARNING, CRITICAL)

3. generate_statistics - Analyze patterns and detect anomalies

4. extract_critical_logs - Extract only critical issues

5. detect_anomalies - Find unusual activity and spikes

6. generate_report - Create formatted HTML/text reports

🚀 Quick Start

1. Clone Repository

git clone https://github.com/BBashh/log_file_mcp.git
cd log_file_mcp

2. Set Up Virtual Environment

# Windows
python -m venv venv
venv\Scripts\activate

# Mac/Linux
python3 -m venv venv
source venv/bin/activate

3. Install Dependencies

pip install -r requirements.txt

4. Test Locally

python mcp_server.py
# Should show: "Log Analyzer MCP Server starting..."
# Press Ctrl+C to stop

5. Configure Cursor IDE

Update your mcp.json file:

Location: C:\Users\[YourUsername]\AppData\Roaming\Cursor\User\settings\mcp.json

{
  "mcpServers": {
    "log-analyzer": {
      "command": "[FULL_PATH_TO_VENV]\\venv\\Scripts\\python.exe",
      "args": ["[FULL_PATH_TO_PROJECT]\\mcp_server.py"]
    }
  }
}

6. Use in Cursor

• Restart Cursor IDE

• Look for green dot in Tools & MCP

• Ask questions in Cursor chat:

  • "Read logs from sample_logs"

  • "How many errors are there?"

  • "Generate a report"

🔧 Configuration

Edit `config.py` to customize:

DEFAULT_LOG_DIR = "C:\\path\\to\\logs"  # Your log directory
LOG_LEVELS = ["CRITICAL", "ERROR", "WARNING", "INFO", "DEBUG"]
STATS_CONFIG = {
    "min_line_length": 5,
    "max_line_length": 1000,
    "rare_threshold": 0.01,
}

📝 Usage Examples

In Cursor Chat:

NOTE: If the chat is giving responses without using the mcp tool, then explicitly mention in the chat to use the xyz tool.

Ex:

Read logs:

Read the logs from sample_logs and show me the last 10 lines using the read_logs mcp tool

Count by type:

How many INFO, WARNING, and ERROR logs are there using `xyz` tool?

Get statistics:

Analyze the logs for patterns and anomalies using `xyz` tool

Extract critical:

Show me all critical logs only using `xyz` tool

Generate report:

Create a comprehensive analysis report of the logs using `xyz` tool.

Ex:

🏛️ Technical Details

Technologies

• FastMCP - MCP server framework

• Python 3.x - Programming language

• Cursor IDE - IDE integration

• JSON-RPC - Communication protocol

Design Patterns

• MVC Pattern - Separation of concerns

• Factory Pattern - Tool creation

• Observer Pattern - Event handling

• Singleton Pattern - Config management

Key Features

• Modular architecture (easy to extend)

• Error handling and validation

• Configuration management

• Report generation (HTML/Text)

• Pattern detection

• Anomaly detection

🔐 Security Notes

• Server runs locally only (no internet exposure)

• Log files processed in memory

• No data stored externally

• Compatible with confidential logs

🐛 Troubleshooting

Green dot not showing?

• Check mcp.json has correct paths

• Verify venv Python path is correct

• Restart Cursor completely

"Module not found" error?

• Make sure venv is activated

• Run: pip install -r requirements.txt

Tools not responding?

• Verify server running: python mcp_server.py

• Check sample_logs/sample.log exists

• Restart Cursor and try again