I used to make them in Figma. Every device size laid out by hand, every caption typed into a text layer, every screenshot dragged into a frame and nudged until it sat right. Then exported, renamed, and uploaded one at a time. For one app, in one language, that is an evening. Add a second language and it doubles. Change a single caption and you do the whole dance again.

For a studio of one, that math is brutal. The constellation holds five shipped apps, and a wording tweak across the catalogue could quietly eat a weekend. Somewhere between the fourth export and the fortieth rename I stopped being annoyed and started taking notes, because irritation that specific is usually a spec in disguise.

What ShotsKit is

ShotsKit is a native Mac app that treats a screenshot set the way Xcode treats a build: one source, many targets.

You compose a screenshot once. Frame, caption, background. Every device size follows from that one layout, so there is nothing to redo when Apple wants another resolution. Versions are cheap, so trying a different angle for a feature costs minutes, not another evening.

Then come the two parts I actually built it for.

Localization is one tap. Pick the locales you ship in and ShotsKit produces the full pack for each of them: captions translated, layouts intact, every size present. The difference between shipping in one language and shipping in eight stops being a cost.

And when a set is ready, it goes straight to App Store Connect. No browser tabs, no dragging forty files into forty slots. ShotsKit uploads the new screenshots and replaces the old ones, and the release is actually, finally done.

Its first customers live upstairs

Like everything here, ShotsKit exists because I needed it. Its first customers are the five shipped stars in this constellation, and every screenshot they wear next will come out of it. That is the quiet advantage of building tools for yourself: the roadmap is just the friction you hit this week.

Why "protostar"

On the homepage, ShotsKit is charted as DSA-06 and classed as a protostar: a star still gathering mass, not yet ignited. It has not reached the Mac App Store yet. It will. When it does, these Field Notes will carry its first light, and if you want to know the exact moment that happens, the subscribe box below does one job only.

When you spend your days building apps, the hours you would have spent learning are the first thing to disappear. So I did the thing I always do when something annoys me enough: I built my way out of it.

Built to buy back time

Whisperer AI started as a tool for one listener: me. Hand it a recording, a video, a podcast, and it hands back a clean transcript, a summary, and the parts that actually matter. Not a vague list of bullet points, but enough that I could skip the three hours and still hold a real conversation about what was said. If I had a question about a recording, I could just ask it, and get an answer drawn straight from the source.

It worked. I used it constantly. The watch-later folder finally started to move, and the time it gave back went straight into building the next thing.

From a tool into a library

Once it was part of my day, the gaps became obvious. A summary is useful in the moment, but I wanted to keep things. So I added PDF export, so any transcript or summary could leave the app as a clean document. Then folders, so all of it had somewhere to live.

That small change turned Whisperer from a one-off tool into a library. People started keeping separate shelves: one for meetings, one for podcasts, one for YouTube, one for university lectures. The app I built to get through a single podcast became the place people kept a whole semester of classes.

The reviews I did not want to read

Here is the harder part to write. Whisperer worked for me, which meant I slowly stopped seeing it clearly. The bugs that mattered were not on my device or in the way I used it, so I never hit them. My users did.

The way I found most of them was the way no developer enjoys: a one or two star review. It stings, it pulls the rating down, and there is no reply that fixes the number.

But every one of those reviews was a map. Each frustrated note pointed at something real I could not see from the inside, and once I could see it, I could fix it. The rating is slow to recover. The app is better because of the people who took the time to complain.

What it taught me

Building for yourself is the fastest way to start and the easiest way to go blind. You stop being a fair test the moment the thing works for you. Whisperer reminded me that shipping is not the end of listening. It is the start of it.

The watch-later folder still fills up. The difference now is that I get to the bottom of it, and so do a lot of other people.

It is the first star in the constellation, DSA-01, and the one I have spent the most time on. That is not an accident. When you build a tool to fix your own problem, you never really stop using it, which means you never stop noticing what is still wrong with it.

The trap of doing too much

The first version was good. People liked it. But it could still feel overwhelming, and overwhelm was the one thing it was supposed to remove. A planner that leaves an ADHD brain feeling behind has failed at the only job that matters.

So a lot of the work since has been subtraction as much as addition. Closing the distance between "I have a vague pile of things to do" and "here is my day."

Planning in seconds

The version I am proudest of is the one running now. Simple has an AI assistant that plans a day in seconds. You hand it the mess in your head and it sorts the priorities, decides what belongs on the visual timeline and what is just a quick todo, and lays the whole thing out. If typing is too much in the moment, you can speak instead, and it understands.

That last part matters more than it sounds. The hardest moment for an ADHD brain is often the blank page, the gap between intention and the first step. Shrinking that gap is the entire point of the app.

Where it is now

As of version 2.3, Apple Calendar and Reminders sync automatically with the timeline and task list, so Simple sits alongside the rest of your day instead of competing with it. You can keep multiple lists so one endless list never becomes its own source of dread. The full run of changes lives on the App Store; I won't turn this into a changelog.

Why I keep going

I am proud of Simple in a way that is hard to explain about software. It helps me on the ordinary days when planning feels impossible. And somewhere along the way it started helping thousands of other people with the same kind of brain. Building the thing you needed, and then watching it work for strangers, is the best part of this job.

Next stop, the Mac

You can already run Simple on a Mac today, but it is the mobile version wearing a desktop. I am building a true native Mac app for it, made for the bigger screen and the way you actually work there. More on that when it is ready.

This is first light for deepskyapps.com.

A studio of one, charted as a constellation

DeepSkyApps is one person. Me. Rather than a wall of app icons and download badges, I wanted the portfolio to feel like what it actually is to me: a small set of fixed points I keep coming back to, each one made on purpose. So the apps are a constellation. Every one is a star with a catalog number, DSA-01 through DSA-06, the way real deep-sky objects get logged before anyone gives them a nicer name.

The line under the title says the quiet part out loud: every star here ships. No vapourware in the sky. If a star has a number, you can download it today. Or, for the newest one still warming up, watch it come online.

Small constellation, full attention

A studio of one changes what the software feels like, and that's the whole point.

There's no growth team inventing reasons to send you notifications, and no meeting where the feature you rely on gets cut. When you leave a review or email me, the person reading it is the person who wrote the code.

Six apps is not many. That's deliberate. I would rather keep a short list of things working well than a long list working badly.

How it's built, for the curious

For the developers who'll inevitably open View Source: the whole site is one HTML file. No framework, no build step, no third-party tracking. It went through fourteen design passes and landed at roughly 34 KB.

A few details I'm fond of:

• The starfield is a single <canvas>: a couple hundred stars, each twinkling on its own clock, drifting and wrapping at the edges.
• The nebulae are SVG feTurbulence noise, tinted through radial gradients borrowed from real Hubble palettes, then masked down to soft clouds.
• The only colour on the entire site lives inside the DEEPSKYAPPS wordmark. Hover it and a spectral wave runs through the letters. Everything else is black, white, and restraint.

None of that is necessary. All of it is the point: the site should feel like the apps do.

What Field Notes will be

Field Notes is the studio log. Expect build notes, App Store experiments (what moved the needle and what quietly flopped), release post-mortems, and the occasional honest account of running everything solo. Written by the same hands that ship the apps, which means no ghostwriters and no content calendar. When there's something worth writing down, it goes here.

That's the instrument check. The sky's open. Pick a star.