$0 to start7 services· 4 auto-onboard

This is what SPARKi runs on

Anyone curious about SPARKi's own production stack

The exact stack powering sparki.cleff.io and sparki-mcp.cleff.io in production

This isn't a curated recommendation — it's SPARKi's actual infrastructure. If you want credibility that these services work together, here's proof: this stack is running SPARKi right now.

What's in the box

Every service has a transparent affiliate disclosure. Most have none.

Vercel

✓ wizard

Web frontend hosting (sparki.cleff.io)

Free tier: 100GB bandwidth/mo, unlimited builds

Clerk

manual

Authentication on the chat UI

Free tier: 10,000 monthly active users

GitHub

✓ wizard

Source control for all SPARKi repos

Free tier: Unlimited private repos + Actions minutes for OSS

Sentry

✓ wizard

Error tracking across sparki-web, sparky, pico

Free tier: 5,000 errors/month

Google Analytics 4

manual

Web traffic analytics

Free tier: 25K requests/day to the Data API

AWS EC2

manual

Self-hosted SPARKy MCP server (t4g.medium on sparki-mcp.cleff.io)

Free tier: 750 hours/mo of t4g.micro for 12 months (this project uses t4g.medium → paid)

Pico

✓ wizard

Uptime monitoring (self-built, monitoring SPARKi itself + every project)

Free tier: Free (SPARKi's own tool)

How to set it up

4 of 7 services are onboarded automatically by SPARKi's wizard. The rest you set up by following the steps below.

  1. 1This package is documentation-only — SPARKi already runs on this stack
  2. 2Use it as reference for credible '$0-ish to start' web infrastructure

Ready? SPARKi can onboard most of this for you.

Open SPARKi and run /sparki start sparki-stackfor <your-project> in a direct message. SPARKi will walk you through 4 services one at a time, verify each API key live, and write the config automatically. 3 services (Clerk, Google Analytics 4, AWS EC2) still need manual setup via vendor dashboards.