Getting Started | vibedecoding

Prerequisites

You'll need the following before getting started:

Bun runtime
Supabase account (free tier works)
AWS account with Terraform installed (for Kafka infrastructure)
OpenAI API key (for the enricher agent)

Clone the Repository

git clone https://github.com/irbull/vibedecoding.git
cd vibedecoding/stream-agents
bun install

Configure Environment

Copy .env.example to .env
Add your Supabase project URL and service role key
Add your OpenAI API key
Kafka credentials come after infrastructure setup (next step)

Deploy Database Schema

In your Supabase dashboard, open the SQL Editor and run the contents of schema.sql. This creates the lifestream schema with all required tables.

Set Up Kafka

The system uses Kafka for durable event transport. The cheapest option is a single EC2 instance running Kafka in KRaft mode (~$15/month):

cd infrastructure/ec2-kafka
terraform init
terraform apply

After Terraform completes, copy the broker address from the output and add it to your .env file. See the repo README for detailed setup instructions and the MSK alternative.

Start the System

You'll need multiple terminal windows. Start each component:

# Terminal 1: Initialize and seed
bun run kafka:init        # Create Kafka topics
bun run seed              # Add test data

# Terminal 2: Publisher (DB → Kafka)
bun run kafka:publish

# Terminal 3: Fetcher agent
bun run agent:fetcher

# Terminal 4: Enricher agent
bun run agent:enricher

# Terminal 5: Materializer (Kafka → DB)
bun run kafka:materialize

Once running, any new links added to the database will flow through the pipeline: fetched, enriched with AI-generated tags and summaries, and materialized back to queryable state.

Next Steps

Read the Start Here guide series to understand the architecture
Check the repo README for troubleshooting
Explore the Stream to see example outputs