AI SDK Slackbot

An AI-powered chatbot for Slack powered by the AI SDK by Vercel.

Features

• Integrates with Slack's API for easy Slack communication
• Use any LLM with the AI SDK (easily switch between providers)
• Works both with app mentions and as an assistant in direct messages
• Maintains conversation context within both threads and direct messages
• Built-in tools for enhanced capabilities:
  • Real-time weather lookup
  • Web search (powered by Exa)
• Easily extensible architecture to add custom tools (e.g., knowledge search)

Prerequisites

• Node.js 18+ installed
• Slack workspace with admin privileges
• OpenAI API key
• Exa API key (for web search functionality)
• A server or hosting platform (e.g., Vercel) to deploy the bot

Setup

1. Install Dependencies

npm install
# or
pnpm install

2. Create a Slack App

1. Go to https://api.slack.com/apps and click "Create New App"
2. Choose "From scratch" and give your app a name
3. Select your workspace

3. Configure Slack App Settings

• Under "App Credentials", note down your "Signing Secret"

• Add the following Bot Token Scopes:

  • app_mentions:read
  • assistant:write
  • chat:write
  • im:history
  • im:read
  • im:write

• Install the app to your workspace and note down the "Bot User OAuth Token"

4. Set Environment Variables

Create a .env file in the root of your project with the following:

# Slack Credentials
SLACK_BOT_TOKEN=xoxb-your-bot-token
SLACK_SIGNING_SECRET=your-signing-secret

# OpenAI Credentials
OPENAI_API_KEY=your-openai-api-key

# Exa API Key (for web search functionality)
EXA_API_KEY=your-exa-api-key

Replace the placeholder values with your actual tokens.

5. Deploy your app

• If building locally, follow steps in the Local Development section to tunnel your local environment and then copy the tunnel URL.
• If deploying to Vercel, follow the instructions in the Production Deployment section and copy your deployment URL.

6. Update your Slack App configuration:

Go to your Slack App settings

• Select your app
• Go to "Event Subscriptions"
• Enable Events
• Set the Request URL to either your local URL or your deployment URL: (e.g. https://your-app.vercel.app/api/events)
• Save Changes
• Under "Subscribe to bot events", add:
  • app_mention
  • assistant_thread_started
  • message:im

Remember to include /api/events in the Request URL.

Local Development

Use the Vercel CLI and untun to test out this project locally:

pnpm i -g vercel
pnpm vercel dev --listen 3000 --yes

npx untun@latest tunnel http://localhost:3000

Make sure to modify the subscription URL to the untun URL.

Note: you may encounter issues locally with waitUntil. This is being investigated.

Production Deployment

Deploying to Vercel

1. Push your code to a GitHub repository

2. Deploy to Vercel:

   • Go to vercel.com
   • Create New Project
   • Import your GitHub repository

3. Add your environment variables in the Vercel project settings:

   • SLACK_BOT_TOKEN
   • SLACK_SIGNING_SECRET
   • OPENAI_API_KEY
   • EXA_API_KEY

4. After deployment, Vercel will provide you with a production URL

5. Update your Slack App configuration:

   • Go to your Slack App settings
   • Select your app
   • Go to "Event Subscriptions"
   • Enable Events
   • Set the Request URL to: https://your-app.vercel.app/api/events
   • Save Changes
   • Under "Subscribe to bot events", add:
     • app_mention
     • assistant_thread_started
     • message:im

Usage

The bot will respond to:

1. Direct messages - Send a DM to your bot
2. Mentions - Mention your bot in a channel using @YourBotName

The bot maintains context within both threads and direct messages, so it can follow along with the conversation.

Available Tools

1. Weather Tool: The bot can fetch real-time weather information for any location.

   • Example: "What's the weather like in London right now?"

2. Web Search: The bot can search the web for up-to-date information using Exa.

   • Example: "Search for the latest news about AI technology"
   • You can also specify a domain: "Search for the latest sports news on bbc.com"

Extending with New Tools

The chatbot is built with an extensible architecture using the AI SDK's tool system. You can easily add new tools such as:

• Knowledge base search
• Database queries
• Custom API integrations
• Company documentation search

To add a new tool, extend the tools object in the lib/ai.ts file following the existing pattern.

You can also disable any of the existing tools by removing the tool in the lib/ai.ts file.

License

MIT