How to Build a Slack Agent with the Slack Agent Skill on Vercel

TL;DR

• Install the Slack Agent Skill, run the wizard with your coding agent, and choose the type of agent you want.
• Follow the five-stage setup process using the Slack Agent Template to go from idea to deployment on Vercel.
• Utilize Workflow DevKit for complex interactions and AI Gateway for versatile model integration.

Prerequisites

Before beginning, ensure you have the following:

• A Slack Workspace where you have permission to add and manage apps.
• A Vercel account for deploying your application.
• Access to a coding agent like Claude Code or OpenCode.
• Basic understanding of Slack API and Git.

Step-by-Step Instructions

1. Install the Slack Agent Skill

1. Access the Skill Directory: Navigate to your coding agent’s skill directory.
2. Install the Skill: Use the following command in the coding agent CLI:

   install skill slack-agent-skill
   

2. Run the Wizard

1. Launch the Wizard: Run the skill using your agent:

   run skill slack-agent-skill
   

2. Define Your Agent: Respond to prompts about your agent's purpose (e.g., "support agent" or "standup bot").

3. Follow the Five-Stage Setup

Stage 1: Project Setup

1. Choose LLM Provider: Select the large language model (LLM) provider for your agent.
2. Scaffold the Project: Your agent will create a new project using the Slack Agent Template.

Stage 2: Slack App Creation

1. Manifest Configuration: Customize manifest.json with app attributes like name and description.
2. Console Guidance: Follow the steps in the Slack API console to create and install the app in your workspace.

Stage 3: Environment Configuration

1. Set Up Secrets: Configure your signing secret and bot token.
2. API Keys: Input any necessary API keys for additional integrations.

Stage 4: Local Testing

1. Dev Server: Start the development server to test real-time bot responses.

   npm start # or equivalent command
   

2. Test Interactions: Send messages to your bot on Slack to see immediate responses.

Stage 5: Production Deployment

1. Deploy to Vercel: Use Vercel CLI to deploy your app:

   vercel --prod
   

2. Environment Setup: Configure any environment variables required for your production app.

Tips and Best Practices

• Iterative Testing: Regularly test each component in Slack before moving to the next setup stage.
• Utilize Templates: Leverage provided templates to avoid common configuration errors.
• Monitor Feedback: Continually gather user feedback to refine your agent's responses and efficiency.

Common Issues

• OAuth Scope Errors: Ensure all required OAuth scopes are set in Slack console.
• Deployment Failures: Double-check Vercel environment variables and logs for deployment issues.
• Response Latency: If response times exceed Slack’s 3-second window, streamline your webhook handling logic.

Next Steps

After successfully deploying your Slack agent:

• Extend Features: Explore adding custom tools by instructing your coding agent to integrate with internal systems.
• Learn Advanced Concepts: Enroll in the Vercel Academy Slack Agents course for comprehensive insights.
• Continuous Integration: Set up automated testing and staging deployments with Vercel’s preview features.

By following this guide, you can efficiently deploy a Slack agent using the Slack Agent Skill and Vercel, customizing it to fit your specific business needs while maintaining efficiency and reliability.