@ai-agent/cli

CLI tool for scaffolding AI Agent projects with the @ai-agent/core framework.

Installation

npm install -g @ai-agent/cli

Or use directly with npx:

npx @ai-agent/cli my-agent-project

Usage

Interactive Mode

Run the CLI without arguments for an interactive setup:

create-ai-agent

You'll be prompted to choose:

• Project name
• Template type (chatbot, Q&A bot, task automation)
• Storage backend (Prisma, in-memory)
• LLM provider (OpenAI, Anthropic, custom)

Command Line Options

create-ai-agent [project-name] [options]

Options:

• -t, --template <template> - Template to use

  • chatbot-prisma - Chatbot with Prisma database storage
  • chatbot-memory - Chatbot with in-memory storage
  • qa-bot - Q&A bot with knowledge base
  • task-automation - Task automation with custom tools

• -s, --storage <storage> - Storage backend

  • prisma - Prisma ORM (PostgreSQL/MySQL/SQLite)
  • memory - In-memory storage (development)

• --skip-install - Skip npm install

• --skip-git - Skip git initialization

Examples:

# Create a chatbot with Prisma storage
create-ai-agent my-chatbot -t chatbot-prisma -s prisma

# Create a Q&A bot with in-memory storage
create-ai-agent my-qa-bot -t qa-bot -s memory

# Create a project without installing dependencies
create-ai-agent my-agent --skip-install

Templates

Chatbot with Prisma

Production-ready chatbot with database persistence using Prisma ORM.

Features:

• Conversation history stored in database
• Support for PostgreSQL, MySQL, or SQLite
• Session management
• Ready for production deployment

Chatbot with Memory

Simple chatbot with in-memory storage for development and testing.

Features:

• Fast setup with no database required
• Perfect for development and prototyping
• Easy to understand code structure

Q&A Bot

Knowledge base bot with RAG (Retrieval-Augmented Generation).

Features:

• Load documents from markdown files
• Semantic search over knowledge base
• Answer questions based on your documents
• Extensible knowledge management

Task Automation

Agent with custom tools for task automation.

Features:

• Example calculator tool included
• Easy to add custom tools
• Tool calling and execution
• Workflow automation

Project Structure

Generated projects follow this structure:

my-agent-project/
├── src/
│   ├── index.ts          # Main application entry
│   └── tools/            # Custom tools (task-automation)
├── prisma/               # Database schema (Prisma projects)
│   └── schema.prisma
├── knowledge/            # Knowledge base (Q&A bot)
│   └── example.md
├── .env.example          # Environment variables template
├── .gitignore
├── package.json
├── tsconfig.json
└── README.md

After Creation

1. Configure environment variables:

   cd my-agent-project
   cp .env.example .env
   # Edit .env and add your API keys

2. Set up database (Prisma projects only):

   npm run db:migrate

3. Run the project:

   npm run dev

Requirements

• Node.js >= 18.0.0
• npm or pnpm

Documentation

• AI Agent Core Documentation
• API Reference
• Usage Guide

License

MIT