Env-ai NPM | npm.io

env-ai - AI Assistant for Your Local Environment

env-ai is an intelligent assistant tool for your terminal, designed to help you with tasks like documentation, performance optimization, refactoring, and more, using custom commands.

!IMPORTANT
env-ai needs the Ollama technology to work. Make sure you have it installed before using this CLI.

📑 Index

🌟 Features

💬 Real-Time Chat: Interact with the AI assistant directly from your terminal.
🗃️ Robust context: Accepts URL and local file inputs so the wizard has more precise context of what to do.
🎨 Extensive Customization: Configure themes, inputs and outputs to your liking.
📄 Configuration File Support: Make your chat systems portable. Compatible with multiple formats, including .mjs, .js, .json, .yml, .yaml, .toml, and .tml.
✈️ Portable: Create a chat system, save it in a configuration file and use it for other projects.
🌐 Multiple Environments:
- 📦 JavaScript Library: Easily integrable into your projects.
- 💻 Command Line Interface (CLI): Works in:
  - 🟢 Node.js
  - 🦕 Deno
  - 🍞 Bun
- 🚀 Binary: Available for all operating systems and architectures via GitHub Releases.

📦 Installation

Install the CLI or add it as a dependency to your project:

## npm
npm install env-ai
## pnpm
pnpm add env-ai
## yarn
yarn add env-ai
# bun
bun add env-ai
# deno
deno add env-ai

global Installation

## npm
npm install -g env-ai
## pnpm
pnpm add -g env-ai
## yarn
yarn global add env-ai

without Installation

## npm
npx env-ai
## pnpm
pnpx env-ai

📖 Using the CLI

The env-ai CLI allows you to easily interact with the AI assistant. Here are some useful commands and options:

Main Commands

# Start a chat with the AI assistant
env-ai chat

Options

-i, --input - Path patterns or URLs to be processed. (array)
-m, --model - Name of the Ollama model to use. (string)
-p, --prompt - Fist prompt to generate a response (text, path or url). (string)
-s, --system - Custom system text or path. (string)
-t, --theme - Topic of conversation (custom, explain, docs, fix, performance, refactor, tests). (string)
-o, --output - Output path for the generated response. (string)
--overwrite - Behavior control if the output file exists (always, ask, last). (boolean)
--single - Get only one response. (boolean)
-c, --config - Path to config file. Files supported: .mjs|.js|.json|.yml|.yaml|.toml|.tml. (string)
--debug - Debug mode. (boolean)
-h, --help - Show help. (boolean)
-v, --version - Show version number. (boolean)

📚 Using the Library

env-ai can also be integrated as a library into your JavaScript or TypeScript project.

Import Example

import { run } from 'env-ai';

run({
    input: ['./src/**', 'https://example.com'],
    theme: 'docs',
    output: 'README.md',
});

Defined Configuration

Use defineConfig to define a reusable configuration:

import { defineConfig } from 'env-ai';

export default defineConfig({
    input: ['./src/**', 'https://example.com'],
    theme: 'docs',
    output: 'README.md',
});

🔍 Examples

You can see more examples here.

CLI

env-ai chat -i "./src/**" -t "docs" -o "output.md"

Library

import { run } from 'env-ai';

run({
    input: ['./src/**', 'https://example.com'],
    theme: 'docs',
    output: 'README.md',
});

CLI With `js` config file

env-ai chat --config dovenv.config.js

import { defineConfig } from 'env-ai';

export default defineConfig({
    model: 'llama3.2:latest',
    input: ['./src/**', 'https://example.com'],
    theme: 'docs',
    output: 'README.md',
});

CLI With `json` config file

env-ai chat --config dovenv.config.json

{
    "theme": "custom",
    "system": "./your-system-content.txt"
}

CLI With `toml` config file

env-ai chat --config documentation-context.toml

theme = "docs"
input = ["./docs", "./src"]
system = """
You are a helpful assistant explaining how to use the provided code library and provide detailed documentation.
The content for the following code library:

{{content}}
"""

CLI With `yaml` config file

env-ai chat --config dovenv.config.yaml

theme: custom
system: ./your-system-content.md

CLI - Input with excluded files

It includes all SRC files that are at the upper level.

env-ai chat -i 'src/*' '!src/../*' -t 'docs' -o 'output.md'

👨‍💻 Development

env-ai is an open-source project and its development is open to anyone who wants to participate.

☕ Donate

Help us to develop more interesting things.

📜 License

This software is licensed with GPL-3.0.

🐦 About us

PigeonPosse is a ✨ code development collective ✨ focused on creating practical and interesting tools that help developers and users enjoy a more agile and comfortable experience. Our projects cover various programming sectors and we do not have a thematic limitation in terms of projects.