Skip to content

ai Command

AI 对话

Overview

启动一个基于配置的 AI 大模型交互式对话会话

Key Features

  • Multi-Model Support
  • OpenAI (GPT-3.5/GPT-4)
  • DeepSeek
  • More models being integrated...
  • Streaming Output
  • Real-time AI response display
  • Typewriter effect support
  • Intelligent Context
  • Automatic conversation history management
  • Multi-turn dialogue support
  • Agent System
  • Built-in professional agents
  • Custom agent support
  • Scenario-based conversation capabilities

Usage

wn ai [flags]

Command Parameters

Parameter Description Default Value Example
--model Specify the AI model to use gpt-3.5-turbo --model=gpt-4
--agent Specify the agent to use default --agent=git-diff
--context Enable context memory true --context=false
--stream Enable streaming output true --stream=false

Use Cases

1. Code Development Assistant

# Get code review suggestions
wn ai "Please review the security of this code" --agent=code-review

# Get refactoring suggestions
wn ai "How can I improve this function's performance" --agent=refactor

2. Git Operation Assistant

# Analyze Git differences
wn ai "Explain the impact of these code changes" --agent=git-diff

# Generate commit message
wn ai "Help me generate a standardized commit message" --agent=git

3. Technical Documentation Translation

# Translate technical documentation
wn ai "Please translate this documentation to English, maintaining accurate technical terms" --agent=translate

Environment Configuration

API Key Configuration

API keys need to be configured before use:

  1. OpenAI API Key
export OPENAI_API_KEY="your-api-key"
  1. DeepSeek API Key
export DEEPSEEK_API_KEY="your-api-key"

Proxy Settings (Optional)

If you need to use a proxy, you can configure:

export HTTP_PROXY="http://127.0.0.1:7890"
export HTTPS_PROXY="http://127.0.0.1:7890"

Advanced Features

Custom Agents

You can create your own specialized agents for specific scenarios:

  1. Create a new markdown file in the agent/agents/ directory
  2. Define the agent's role, tasks, and expertise
  3. Use the --agent parameter to specify the newly created agent

Context Management

  • Use --context=false to enable context-free mode
  • Use the "clear context" command to reset conversation in long dialogues
  • Support referencing historical messages in replies

Common Issues

Q1: How to switch between different models?

A1: Use the --model parameter to specify the desired model, for example: --model=gpt-4 or --model=deepseek-chat

Q2: Why is the response slow?

A2: Possible reasons: - Network connection issues - API rate limiting - Model processing time

Suggestions: - Check network connection - Configure appropriate proxy - Choose faster responding models

Q3: How to handle API key expiration?

A3: Update the API key in environment variables:

export OPENAI_API_KEY="your-new-api-key"

Best Practices

  1. Choose Appropriate Agents
  2. Use specialized code agents for code-related questions
  3. Use translation agents for translation tasks

  4. Use architecture agents for design tasks

  5. Effectively Use Context

  6. Keep related questions in the same session
  7. Clear irrelevant context timely

  8. Use context-free mode when appropriate

  9. Optimize Output Results

  10. Use clear and specific questions
  11. Properly use code blocks and examples
  12. Choose suitable models based on requirements