/log-ai-analyzer

A PowerShell-based tool for parsing, diagnosing, and optionally AI-analyzing Intune/Autopilot deployment logs.

Primary LanguagePowerShellMIT LicenseMIT

LogAI Analyzer — Universal AI-Enhanced Log Parser (PowerShell Tool)

Version: 6.0

Author: MD Faysal Mahmud (faysaliteng@gmail.com)

Repository: https://github.com/Blindsinner/log-ai-analyzer

Enhanced With: Google Gemini, OpenAI, Azure OpenAI, unified offline/online error analysis, multi-format support, and responsive HTML exports.

Another Log Analyzer Web Verison: AI-Powered-Log-Analyzer-Ultimate GUI (WEB)

📖 Overview

LogAI Analyzer is a powerful PowerShell tool designed to parse and diagnose errors from Microsoft Intune, Autopilot, and other Windows log sources. It intelligently combines a local error database (error_db.json) with AI-powered analysis from Google Gemini, OpenAI, and Azure OpenAI to deliver comprehensive diagnostics.

Key Features:

  • Universal File Support: Natively handles .zip, .xlsx, .xls, .evtx, .etl, .log, .txt, .html, .xml, and more.

  • Recursive Archive Extraction: Automatically processes nested archives to find all relevant logs.

  • Hybrid Analysis:

    • Offline: Fast lookups using the local error_db.json.
    • Online: AI-driven diagnostics for unrecognized or complex errors.

  • Custom Keyword Detection: Define your own search tokens in the errorcloud.txt file.

  • Responsive Reports: Generates professional, SEO-friendly HTML and plain-text reports.

  • Cross-Platform: Compatible with Windows PowerShell 5.1+ and PowerShell Core (macOS/Linux).

📋 Table of Contents

  1. Prerequisites
  2. Installation & Cross-Platform Setup
  3. Configuration
  4. File & Folder Structure
  5. Usage Examples
  6. Main Menu Options
  7. Core Logic
  8. Output
  9. Troubleshooting
  10. Contributing & License

🛠️ Prerequisites

  • PowerShell:

    • Windows: PowerShell 5.1 or later.
    • macOS/Linux: PowerShell Core v7.2+ (pwsh).

  • PowerShell Module: The ImportExcel module is required for .xlsx and .xls file support.

  • Permissions: Administrator privileges are required on Windows to analyze .evtx and .etl files.

  • Internet Connection: Required for AI-based analysis and optional online searches.

  • Required Files: The LogAnalyzer.ps1, error_db.json, and errorcloud.txt files must be located in the same directory.

⚙️ Installation & Cross-Platform Setup

Windows (Run in an Administrator PowerShell)

  1. Run as Administrator: Search for “PowerShell”, right-click it, and select Run as administrator.

  2. Verify Git Installation:

    git --version

  3. Get the Code:

    • With Git (Recommended):

      git clone https://github.com/Blindsinner/log-ai-analyzer.git
cd log-ai-analyzer

    • Without Git:

      1. Navigate to https://github.com/Blindsinner/log-ai-analyzer.

      2. Click Code → Download ZIP.

      3. Extract the archive to a known location (e.g., C:\Users\YourUser\Downloads\log-ai-analyzer).

      4. In PowerShell, navigate to the directory:

        cd "C:\Users\YourUser\Downloads\log-ai-analyzer"

  4. Install Required Module:

    Install-Module -Name ImportExcel -AcceptLicense -Force

  5. Unblock the Script:

    Unblock-File -Path .\LogAnalyzer.ps1

  6. Run the Analyzer:

    .\LogAnalyzer.ps1

macOS / Linux

  1. Install PowerShell Core (Example for Ubuntu):

    wget -q https://packages.microsoft.com/config/ubuntu/20.04/packages-microsoft-prod.deb
sudo dpkg -i packages-microsoft-prod.deb
sudo apt-get update
sudo apt-get install -y powershell

  2. Clone Repository & Navigate:

    git clone https://github.com/Blindsinner/log-ai-analyzer.git
cd log-ai-analyzer

  3. Install Required Module:

    pwsh -Command "Install-Module -Name ImportExcel -AcceptLicense -Force"

  4. Run the Analyzer:

    pwsh ./LogAnalyzer.ps1

Note: Analysis of .evtx and .etl files is only supported on Windows. All other formats are fully functional across all platforms.

🔧 Configuration

  1. API Keys: The first time you use an AI-powered feature, the script will prompt you to enter and save your API keys. The keys will be stored locally in gemini_key.txt, openai_key.txt, and azure_key.txt.
  2. Offline Database: You can extend the local database by editing error_db.json. Add new entries using the format: { "ErrorCode": "0x12345678", "Message": "A brief description.", "Solution": "A detailed solution." }.
  3. Custom Keywords: Add your own comma-separated keywords to errorcloud.txt (e.g., timeout,access denied,failed to connect) to customize error detection.

📂 File & Folder Structure

log-ai-analyzer/
├── LogAnalyzer.ps1         # Main script
├── error_db.json           # Offline error database
├── errorcloud.txt          # Custom keyword list
├── gemini_key.txt          # Created automatically on first AI run
├── openai_key.txt          # Created automatically on first AI run
├── azure_key.txt           # Created automatically on first AI run
└── Analyzed Results/       # Default directory for output files
    ├── LogAnalysis_YYYYMMDD_HHMMSS.html
    └── AI_Analysis_YYYYMMDD_HHMMSS.txt

🚀 Usage Examples

.\LogAnalyzer.ps1
# Select option 1 from the menu
# Enter the path to your log file, e.g., C:\Logs\example.evtx

🔍 Main Menu Options

=== Universal Log Analyzer Menu (v6.0 Gemini Enhanced) ===
1. Analyze Log File (Offline DB & optional online search)
2. Analyze with AI Only (Directly analyze log with selected AI)
3. Select AI Model (Current: gemini-pro)
4. Manage AI Providers & API Keys
5. Exit
Choose an option (1-5):

🧩 Core Logic

  1. File Identification: The script first determines the file type based on its extension.
  2. Extraction & Parsing: Archives are recursively extracted, and Excel files are imported. Event logs (.evtx), trace logs (.etl), and plain-text logs are parsed accordingly.
  3. Pattern Matching: The content is scanned for hexadecimal error codes, error-related phrases, and custom keywords from errorcloud.txt.
  4. Hybrid Resolution: Detected errors are first looked up in the local error_db.json. If an error is not found, it is sent to the selected AI model for advanced analysis.

📊 Output

  • Console: Real-time progress and results are displayed in neatly bordered sections.
  • Text Files: *.txt files provide a simple summary of findings.
  • HTML Reports: *.html files offer a responsive, card-based layout for easy reading and sharing.
  • All reports are saved to the Analyzed Results/ folder with a timestamp for clear organization.

🛠️ Troubleshooting

  • ImportExcel Module Not Found: Ensure the module is installed by running Install-Module ImportExcel -Force in an administrative PowerShell session.
  • Permission Denied Errors: The script requires elevated (administrator) privileges to read certain system logs (.evtx, .etl). Right-click PowerShell and select "Run as administrator."
  • No Errors Detected: Double-check that the file path is correct and that the logs contain searchable error codes or keywords defined in errorcloud.txt.
  • AI Analysis Fails: Verify that your API keys are correct and that you have a stable internet connection. Check the AI provider's status page for any outages.

🤝 Contributing & License

We welcome contributions! Feel free to open issues, submit pull requests, update the error database, or improve the documentation.

This project is licensed under the MIT License. See the LICENSE file in the repository for details.