AI Analysis for SSH Command Line

• AI Risk Assessment: A new "Risk Level (AI)" column is added to the SSH session details, categorizing executed commands into High, Medium, and Low risks based on AI evaluation.
• Command Categorization: The system intelligently displays the "Type" of commands executed within the SSH session details.
• Interactive "Ask AI" Assistant: An "Ask AI" button is integrated into the SSH historical session details, allowing you to ask the AI questions about the specific session. Note that the AI chat history is not saved if you navigate away from the page.
• Supported Models: The system supports integrations with OpenAI, Ollama, and Qwen models. It is recommended to use pure text models rather than multimodal models to save token.

1. Verify and Import License: Navigate to System >> System License and ensure you import a license that explicitly supports and enables the AI module.
2. Add the AI Model: Go to System >> AI Models and add an available model. You will need to configure the parameters based on the model type:
   • OpenAI: Both the Basic URL and the API Key are required.
     • Deepseek is also compatible to use.
   • Ollama: Only the Basic URL is required; the API Key should be left blank.
   • Qwen: The API Key is required. The Basic URL is optional (if left blank, it automatically defaults to https://dashscope.aliyuncs.com).
   • Important Note: If multiple AI models are enabled simultaneously, the system will use the first record queried from the database. If the AI settings are disabled or the model connection fails, the AI risk levels will be empty and the "Ask AI" button will be unavailable.
3. Test Connectivity: Once the model is added, click the "Test" button on the right side of the list to verify the connection between the PAM system and the AI model.

Store the API Key into the AI Module in RankEZ.