AI Settings

AI functionality is a core feature of Chaterm, providing intelligent operations assistant capabilities. This document will introduce how to use Chaterm's AI features. With proper configuration, you can fully utilize the AI assistant to improve work efficiency.

AI Dialog Modes

Click "AI" in the left menu to open the AI dialog panel. AI dialog modes support:

• Natural Language Interaction: Use natural language to describe tasks, AI will understand your intent
• Multi-Tab Dialogs: Can conduct multiple AI dialogs simultaneously, each dialog is an independent session
• Context Understanding: AI will understand the current connected server environment and context information
• Task Planning: AI can plan complex operations tasks and execute them step by step

Interaction Modes

AI supports three interaction modes:

Chat Mode

• Function: Dialog with AI, learning, brainstorming
• Features: Pure text conversation, similar to ChatGPT
• Use Cases:
  • Knowledge learning and consultation
  • Brainstorming
  • Technical discussions
  • Code review

Note

Terminal operations are not supported in this mode, cannot execute server commands.

Command Mode

• Function: Execute related tasks in the current active terminal
• Features:
  • All input and output are displayed in the terminal interface
  • AI will plan tasks and execute commands
  • Requires user confirmation before execution (no confirmation needed if auto-execute is enabled)
• Use Cases:
  • Command-line operations
  • Local task execution
  • Single terminal task processing

Agent Mode

• Function: Execute command queries on any host, troubleshoot errors, handle tasks, and more
• Features:
  • Operation results are displayed in the AI content area
  • Supports fully automatic execution
  • Can @ multiple hosts for operations
  • AI will adjust strategy based on execution results and continue completing tasks
• Use Cases:
  • Multi-host task processing
  • System problem troubleshooting
  • Automated operations
  • Complex task planning

Create New Dialog

Click "AI" in the left menu to open the AI dialog panel. Click new dialog to configure:

• Interaction Mode: Supports Chat, Command, Agent mode selection
• @Host: Agent mode supports @ multiple hosts for operations
• Model: Can select different models from the model list
• Upload File: Upload files, AI will read file content to understand and process tasks
• Voice Input: Supports user voice input

History Dialogs

The AI dialog panel supports history dialog management:

• View History: Click the menu button in the top left to view all history dialogs
• Load Dialog: Click on history dialog items to load previous dialog content
• Delete Dialog: Can delete unwanted history dialogs
• Dialog Title: System automatically generates dialog titles, can also be manually modified
• Favorite Dialog: Can favorite needed history dialogs

Export Chat

Supports exporting history dialogs for easy saving and sharing.

Model Settings

Select API Provider

Chaterm currently supports the following AI service providers:

• LiteLLM: Supports multiple model providers
• OpenAI: OpenAI compatible interface
• Bedrock: Amazon Bedrock
• DeepSeek: DeepSeek API
• Ollama: Local Ollama service

Fill in Authentication Information

Based on the API provider you selected, you need to fill in the corresponding authentication information:

• OpenAI Compatible: Need to fill in API Key and service endpoint
• Amazon Bedrock: Need to configure AWS credentials
• DeepSeek: Need to fill in API Key
• LiteLLM: Configure according to specific provider
• Ollama: Configure local Ollama service address

Select Model

After filling in authentication information, you can:

• Directly enter model name
• Or select available models from the dropdown list

Verify Configuration

After completing the above settings:

1. Click the "Test Connection" button on the interface
2. System will automatically test the connection
3. If configuration is correct, a success prompt will appear
4. You can now start using AI features

Notes

• Please ensure the authentication information filled in is accurate
• It is recommended to regularly check the validity of API keys
• If encountering connection issues, please check network status and service provider status
• For detailed configuration, please refer to the Model Configuration documentation

Prerequisites

Before starting to use AI features, please ensure you have:

• Completed AI configuration in settings
• Tested that AI connection status is normal
• Selected an appropriate interaction mode

For detailed configuration instructions, please refer to Model Configuration and AI Preferences.