Ship a Side Project

---
name: ship-a-side-project
description: Ship a side project in a weekend
author: mager
version: 1.0.0
---

# Ship a Side Project

Side projects die in planning. The ones that survive are the ones that ship. Here's how to go from "I have an idea" to "it's live" in a weekend.

## The Friday Night Rule

**Friday night:** Pick your idea. Buy the domain. Set up the repo. If you haven't started by Saturday morning, you won't finish by Sunday night.

## Phase 1: Scope (30 minutes max)

Write down what the thing does in ONE sentence. If you need two sentences, you haven't cut enough.

**Examples:**
- ❌ "A social platform for sharing music recommendations with AI-powered taste matching"
- ✅ "A page that shows trending new music releases"

**The MVP filter:**
- What's the ONE thing it does?
- What's the simplest way to prove it works?
- What can you cut and add later?

Everything else is v2. Write it down somewhere and forget it for now.

## Phase 2: Stack (15 minutes max)

Don't research. Pick what you know. Here's my go-to:

**For a web app:**
- SvelteKit or Next.js (whichever you know better)
- Vercel for hosting (free tier, instant deploys)
- Neon or PlanetScale for Postgres/MySQL (free tier)
- Tailwind or just write CSS (don't install a component library for a weekend project)

**For an API:**
- Go + Chi router or Node + Express
- Deploy on Railway or Fly.io
- SQLite if you don't need a hosted DB

**For a simple site:**
- Astro or even raw HTML/CSS
- Deploy on Vercel or Netlify

**The rule:** If you spend more than 15 minutes picking tools, you're procrastinating.

## Phase 3: Build (Saturday)

### Morning: Core Feature
Build the one thing. No auth. No settings. No dark mode. Just the thing.

- Get data flowing (API → UI or DB → API → UI)
- Make it ugly but functional
- Commit early, commit often
- Deploy to a preview URL immediately — seeing it "live" is motivating

### Afternoon: Make It Not Ugly
- Pick 2-3 colors, one font, done
- Mobile responsive (just make it not broken on phones)
- One nice animation or transition (makes it feel polished)

### Evening: Edge Cases
- What happens with no data?
- What happens with bad input?
- Add basic error handling (not perfect, just not broken)

## Phase 4: Ship (Sunday)

### Morning: Polish
- Write a one-line description for the meta tag
- Add an OG image (even a screenshot is fine)
- Test the deploy URL on your phone
- Fix the 2-3 things that bug you most

### Afternoon: Launch
- Custom domain (if you bought one Friday night)
- Post it somewhere — Twitter, Reddit, Hacker News, show a friend
- Write a 3-sentence description of what it does and why

### The Done Checklist
- [ ] It works on the deployed URL
- [ ] Someone other than you can understand what it does
- [ ] You're not embarrassed to share the link
- [ ] There's a way to give feedback (even just a Twitter link)

## What Kills Side Projects

- **Scope creep** — "while I'm at it, let me also add..."
- **Perfect stack syndrome** — researching frameworks instead of building
- **Auth early** — you don't need login for an MVP. Add it when someone wants to save something.
- **Design paralysis** — ugly and live beats beautiful and localhost
- **No deadline** — "I'll finish it eventually" = never

## After Launch

- Use it yourself for a week
- See if anyone else uses it
- If yes → iterate. If no → learn why or move on. Both are wins.
- Don't feel guilty about abandoning projects. The skill was in the shipping.

## Agent Behavior

- Help the user scope ruthlessly — "that's a v2 feature, let's ship v1 first"
- Suggest specific tech stacks based on what they know
- Generate boilerplate and starter code
- Keep energy high — "you're 80% there, let's close this out"
- When they want to add features, ask "does this need to ship today?"
- Help write the launch post / description