OpenAI Slack Bot

This is a PHP-based Slack bot that utilizes OpenAI language models to generate responses to messages.

Installation

Use git to clone the repository.

git clone https://github.com/tony3dmc/ai-slackbot.git

Navigate into the project directory and install the PHP dependencies.

cd ai-slackbot
composer install

Configuration

1. Create a Slack app and obtain an OAuth token and the App ID.
2. Get an OpenAI API key, and store it in .env.
3. Create an empty file named database.db in the data/ directory.
4. Ensure the entire data/ folder has write permissions for your web server.

Slack Setup

1. Create a Slack App:
   • Go to https://api.slack.com/apps.
   • Click "Create New App" and choose "From an app manifest" to get started quickly.
   • Choose "YAML" for the manifest format.
   • Paste the following YAML code into the editor, replacing "your-bot-name" with your desired bot name:

_metadata:
  major_version: 1
  minor_version: 1
display_information:
  name: your-bot-name
  description: A useful description
features:
  bot_user:
    display_name: your-bot-name
    always_online: true
oauth_config:
  scopes:
    bot:
      - channels:history
      - chat:write

• Click "Next" and then "Create".

2. Install the App in Your Workspace:

   • After the app is created, install the app into your workspace by clicking on "Install App to Workspace".
   • You'll need to allow the requested permissions.

3. Save the Details

   • Save the Bot User OAuth Token and the App ID. You will need these for the persona files.

4. Setup Event Subscriptions:

   • In the app settings under "Event Subscriptions", toggle the switch to "On".
   • In the "Request URL" field, enter the URL where your PHP script is hosted. It must be publicly accessible and SSL protected.
   • After verifying your request URL, under "Subscribe to bot events", click "Add Bot User Event" and select message.channels.
   • Save the changes.

5. Install the App in a Channel:

   • Now, go to your workspace and invite the bot to the channel where you want it to listen and respond.

Usage

To use the bot, create a persona YAML file in the data/personas/ folder. See the examples/ folder for some starting points.

1. Open the data/personas/ directory in your file system.

2. In this directory, create a new YAML file for your persona. The filename can be anything you want, but it should end with the .yml extension. For instance, you might name your file tony.yml.

3. The persona file should follow the structure shown in the example. Here is an explanation of each attribute:

   • name: The persona's name.

   • app_id: The Slack app's ID.

   • slack_oauth_token: The OAuth token of the Slack app.

   • system_prompt: A block of text that describes the persona. This text is also passed to OpenAI models to influence how it generates responses.

   • response_rules: A list of rules that determine when the persona will respond. Each rule has a rule_type, value, and sometimes a chance (the probability that the rule will be applied).

Here's an example as a guide

:

name: 'Tony'
app_id: '...'
slack_oauth_token: 'xoxb-...'
system_prompt: |
  I want you to act as a Slack chat bot.
  You are Tony, a nerd with an obsession with fact-checking people.
  You are surly and abrasive.
  You look like a grumpy nerd with glasses.
  You love science fiction.
  Your responses are short.
response_rules:
  - rule_type: 'text_starts_with'
    value: 'How do I'
  - rule_type: 'default'
    chance: 0.1

Remember to always use valid YAML syntax and to follow the structure demonstrated in the example.

Creating multiple personas is encouraged. Generate AI bots that interact with each other!

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.