SlackGPT

A slack integration for OpenAI's Chat GPT on AWS

Features

• Responds to mentions
• Uses chat history
• Customizable system prompt on deployment

## Limitations

• API Gateway's 30 seconds response time => may cause timeouts if the prompt (including chat history) requires longer processing by Chat GPT
• No robust error handling
• Deployment may require manual actions => See Considerations section

Getting Started

This section will guide you through the setup process

Pre-requisites

• AWS account
• OpenAI account
• Slack account

Deployment

This section will walk you through the deployment process

1. Create a Slackbot

Follow the link in the official documentation for instructions. You can use the provided App Manifest to create the app instead of doing it from scratch, or use it as a reference for the manual configuration. However make sure you comment out / omit the event_subscriptions section untill AWS deployment is complete.

2. Install the app to your workspace

3. Generate authorization tokens

   • Generate a Slackbot token from OAuth & Permissions
   • Create a ChatGPT token here

4. Build & Deploy the project

npm run build -s
npm run deploy -s

6. Enable event_subscriptions for the bot using the SlackRequestUrl (see Outputs section after deployment) as request_url.

Follow the instructions of AWS SAM and use your GPT and slack tokens to

Considerations

After deployment some you need to perform the following manual changes on the created resources:

• Give the Http API gateway permissions to invoke lambda (done with AWS CLI)
• Enable TTL for ttl parameter of DynamoDB table