Skip to main content

Examples

Here are some practical examples of how to use AI CLI in your development workflow.

Search for Solutions​

Debugging a TypeScript Error​

ai search "TypeError: Cannot read property 'map' of undefined in React component"

This will:

  1. Search GitHub issues and Stack Overflow for similar errors
  2. Analyze the results with AI
  3. Provide a detailed solution with code examples

Finding Best Practices​

ai search "React hooks best practices useEffect cleanup"

Troubleshooting Build Issues​

ai search "pnpm install error ENOENT: no such file or directory"

Generate Documentation​

Document a React Component​

ai describe src/components/UserProfile.tsx

This will create src/components/UserProfile.md with:

  • Component overview
  • Props documentation
  • Usage examples
  • Key implementation details

Document an API Service​

ai describe src/services/api.ts

Debug Code​

Debug a Complex Error​

ai debug
# Then paste your error:
TypeError: Cannot read property 'map' of undefined
    at UserList (./src/components/UserList.tsx:15:8)
    at renderWithHooks (./node_modules/react-dom/cjs/react-dom.development.js:14803:9)
^D

# Then paste the relevant code:
const UserList = ({ users }) => {
  return users.map(user => <UserCard key={user.id} user={user} />);
};

Collect and Analyze Code​

Analyze a Feature​

# Collect all TypeScript files in the feature directory
ai collect ./src/features/user-management

# This will:
# 1. Create a concatenated file in .ai-cli/user-management.txt
# 2. Generate an analysis in .ai-cli/user-management_analysis.md

Review a Module​

# Collect and analyze a specific module
ai collect ./src/utils

Configuration Examples​

Set Up Multiple AI Models​

# Use Claude for search (faster, good for general queries)
ai config-model search anthropic claude-3-haiku-20240307

# Use GPT-4 for debugging (more detailed analysis)
ai config-model debug openai gpt-4

# Use Claude for documentation (good balance of speed and quality)
ai config-model describe anthropic claude-3-sonnet-20240229

Configure API Keys​

# Set up Anthropic API key
ai config anthropic sk-ant-xxxxx

# Set up GitHub token for search
ai config github ghp_xxxxx

# Verify configuration
ai config-list

Real-World Workflow​

Here's a complete example of using AI CLI in a typical development workflow:

  1. Start a New Feature
# Collect existing code to understand the codebase
ai collect ./src/features
  1. Debug an Issue
# Get AI-powered debugging help
ai debug
# Paste error and code
  1. Search for Solutions
# Find similar implementations
ai search "React form validation with TypeScript"
  1. Document Your Code
# Generate documentation for new components
ai describe src/features/new-feature/components/Form.tsx
  1. Review Changes
# Collect and analyze modified files
ai collect ./src/features/new-feature

Tips and Best Practices​

  1. Use Search with Specific Limits
# Get more detailed results
ai search "React hooks" -g 5 -s 5
  1. Combine Commands
# Generate documentation and then search for improvements
ai describe src/components/DataTable.tsx
ai search "React table component best practices"
  1. Use Interactive Debug Mode
ai debug
# Take your time to paste the complete error stack trace
# Then provide relevant code context
  1. Keep Documentation Up to Date
# After making changes, regenerate documentation
ai describe src/components/UpdatedComponent.tsx