ghee

A simple bot framework for Slack written in ES2016+.

Installation

Via npm

$ npm install ghee

Via git clone

$ npm clone [e-mail]:elliottcarlson/ghee && cd ghee
$ npm install

Usage

ghee is a framework for writing bots, and does not run on it's own. For pre-made bots, please see the Samples section.

If you want to quickly create your first ghee bot, it is recommended to use ghee-boilerplate as it will provide you with everything you need to quickly get a bot up and running.

ES7 + Decorators Bot

ghee comes with a helper function that is intended on being used as a decorator. To use the @ghee decorator, you will need babel-preset-es2015, babel-preset-es2017 and babel-plugin-transform-decorators-legacy. The ghee-boilerplate provides all the files and references needed to quickly get setup to create your own bot.

The ghee helper function will either register the method being decorated directly, or can be passed a parameter to register as the string to respond to. An special parameter of * will cause that method to receive all content and acts as a catch-all method.

A straight-forward bot that will respond to .hello and .goodbye messages in Slack would look like:

import { Ghee, ghee } from 'ghee';

class Bot extends Ghee {
    constructor() {
        super(YOUR_SLACK_API_TOKEN);
    }

    @ghee
    hello() {
        return 'Hello!';
    }

    @ghee('goodbye')
    other_method() {
        return 'Goodbye!';
    }
}

ES6 Bot

If you don't want to use decorators, and want to stick with babel-preset-es2015, you can still use ghee.

The same sample as above, but without a decorator:

import { Ghee, ghee } from 'ghee';

class Bot extends Ghee {
    constructor() {
        super(YOUR_SLACK_API_TOKEN);

        ghee(this, 'hello');
        ghee(this, 'goodbye');
    }

    hello() {
        return 'Hello';
    }

    goodbye() {
        return 'Goodbye!';
    }
}

How it works (quick overview)

ghee abstracts the background comminucation with Slacks RTM and Web API's. By extending the Ghee base class, your Bot will perform all of the connection and routing of requests behind the scenes.

By registering specific methods via the ghee helper (either as a decorator, or directly), ghee will register that methods name as a command that it can respond to.

Your bot will respond to various styles of sending commands. In any room that the bot has been invited to, or via private message, you will be able to trigger a command call using the following syntaxes:

• @-mention command (optional parameters)
• @bot-name command (optional parameters)
• bot-name: command (optional parameters)
• prefix command (optional parameters)

If the command has been registered via the ghee helper, then the method will be called with the following parameters:

• params: an array of individual words that were entered after the command
• from: an object containing the user that issued the command
• channel: an object containing the channel the command was issued in (can indicate direct message as well)
• msg: an object containing the raw message that was received from Slack

Please see the wiki for a more in-depth usage guide.

command's can respond in various ways. In the above examples, we simply return a string - this tells ghee to respond to the command by sending the returned string back to the source - i.e. if it were in a channel, it would respond there, if the request was via direct message, it would respond there. Besides returning a string, ghee can also accept the following return types:

• Attachment: An attachment is a seperate class that is available with ghee - attachments allow you to include additional content with a post, and add styling and fields. For more information on Slack attachments, see their documentation.
• Promise: When a Promise is returned, ghee will act accordingly, and wait for a resolve or reject message to come through. The content of the resolve/reject message should be a string or an attachment that it can process accordingly.

Please see the wiki for more information on Attachments and Promises.

Why "ghee"?

Why not? it's clear to use and smooth as butter.

Samples

• ghee boilerplate
• Sample ghee bot
• JiraBot
• IncidentBot