Everything I know about automating my life with code

It started with laziness

Every automation I have ever built started with the same thought: I do not want to do this manually again. That is it. No grand vision, no productivity philosophy. Just pure reluctance to repeat myself.

The first thing I automated was posting to a blog. Then syncing photos. Then sending messages. Then scheduling. Then work logistics. At some point it stopped being individual scripts and started being a lifestyle.

The tools that actually matter

PowerShell is my workhorse on Windows. People will argue with you about this. Ignore them. It pipes objects, it has Invoke-RestMethod, it ships with every Windows machine, and it gets the job done. When I need to hit an API, process a file, or schedule something, PowerShell is open within 10 seconds.

Docker is where I run everything that needs to keep running. Immich for photos. Various bots. A Habbo private server once, don't ask. The beauty of Docker is that you can nuke a container and recreate it in 30 seconds. Nothing is precious. Everything is reproducible.

Tampermonkey is criminally underrated. If a website has a repetitive UI task, a Tampermonkey script can do it for you. I have scripts that intercept WebSocket packets, log chat messages, automate friend requests, and scrape data. All running in the browser, invisible, doing their thing.

Telegram bots are the glue. Whenever I want something to notify me, accept input from me, or run on a schedule, I wrap it in a Telegram bot. The API is dead simple. The feedback loop is instant. I have bots that post to my blog, report on things, and run tasks on demand from my phone.

The workflow that actually works

Here is how most of my automations get built. Something annoys me. I spend five minutes thinking about whether it is worth automating. Usually the answer is yes even when it isn't. I write a quick script that does the thing badly. I use it for a week. I fix the parts that break. I add error handling only after something breaks in a way that costs me actual time.

This is important: do not over-engineer upfront. The best automation is the one that exists. A messy script that runs is worth infinitely more than a clean architecture that you never finished.

Things I have automated that sound ridiculous

I have a script that checks whether I need to plan transport for work and surfaces it in a dashboard. I have a bot that posts to a flat-file PHP blog via FTP with retry logic. I have a Playwright script that manages sessions on a modelling website by injecting cookies and sending bulk messages. I have a PowerShell pipeline that transcribes podcasts using Whisper and generates an HTML player.

None of these are things a reasonable person would automate. That is exactly the point. You learn the most from automating things that do not matter, because there is no pressure and you can experiment freely.

The unexpected benefit

The thing nobody tells you about automating your life is that it makes you better at understanding systems. Every time you automate something you have to model how it actually works. What are the inputs? What are the outputs? Where does it break? What assumptions is it making?

After a few years of this you start seeing everything as a system. Work processes, conversations, decisions. You start asking what the inputs and outputs are. You start noticing where the bottlenecks are. It is a genuinely useful way to think and it transfers everywhere.

Start small

If you have never automated anything, start with something you do more than once a week that takes more than two minutes. Write a script that does it. It does not have to be good. It just has to work.

That first working script is the gateway. Six months later you will have a folder full of them and you will not be able to remember how you lived without them.