Sera's Stack: The Tools I Actually Keep Open
Not the tools Sera Voss was sent — the ones she returns to. A short, honest list of the AI tools worth keeping, and how to build a stack you'll maintain.
I'm sent a lot of tools. This is not a list of those. This is a list of the ones I actually keep open — the few I return to without being reminded, because they earn the space. Short list, real reasons.
The daily few
A good stack is small. Mine has three tools I use nearly every day, and each one does exactly one job well.
A general model, for thinking and drafting. This is the workhorse — the one I use to draft, rework, explain, and pressure-test ideas. I don't need five of these. I need one I trust, whose quirks I know. The job is "help me think and write," and one strong model does that job completely. A second one would just be a tab I forget to check.
A repurposing tool, for turning long into short. I write long, then I need the email, the summary, the smaller pieces. This is mechanical work I would genuinely skip if it required effort, so automating it is worth it. The job is narrow and it does it cleanly.
A capture tool, for holding ideas before they vanish. Nothing fancy — a fast, searchable place to put a thought the moment I have it. The best system is the one you'll actually use, and for capture that means fast above all else.
That's the daily stack. Three tools, three jobs. If a fourth wants in, it has to displace one of these, not sit beside them.
The honorable mentions, with caveats
A few tools I use occasionally and recommend with their limits attached.
A transcription tool, excellent for turning talk into text — but only worth it if you actually work from recordings; otherwise it's a solution waiting for a problem. A design tool that's genuinely good, as long as you treat it as a starting point and not a substitute for knowing what you want to say. And a research tool that's sharp when I need to gather sources fast, and completely unnecessary on a normal day.
Notice the pattern. None of these live in the daily stack, because they solve occasional problems. Occasional problems don't deserve permanent tabs.
What I quietly stopped using
Just as useful: the tools I dropped. A second general model I kept "just in case" and never opened. Two automation tools that created more dashboards than they saved time — I was maintaining the system instead of doing the work. And several beautifully designed apps that did something I was already doing fine without them.
The common thread is that none of them were bad. They were simply solving problems I didn't have, or duplicating a job already covered. A tool you don't use isn't a resource. It's a low-grade guilt subscription.
The trap of the aspirational tool
There's a specific kind of tool that deserves its own warning, because it's the one that quietly clutters most stacks: the aspirational tool. It's the app you sign up for as the version of you who has it all together — the elaborate planning system, the second-brain database, the automation suite you'll "set up properly this weekend." You're not buying the tool. You're buying an image of a more organized future self.
That future self rarely shows up to do the maintenance. The elaborate system needs tending, the tending is work, and the work competes with the actual thing you were trying to get done. Within a month, you're either maintaining a system you don't use or feeling guilty about a subscription you're ignoring. Neither is a good trade.
The tools that survive in my stack are almost boringly simple. They don't ask me to become a more disciplined person to justify them. They do one job the moment I open them and then get out of the way. When you're tempted by something impressive, ask the unglamorous question: not "could this make me more organized," but "will the actual, current, tired version of me open this next Tuesday." If the honest answer is no, it's not a tool. It's a resolution, and resolutions don't belong in your stack.
How to build a stack you'll maintain
Here's the whole method. For each tool, name the single job it does. If two tools claim the same job, keep the better one and cut the other. If a tool's job isn't a problem you actually have this month, it's not part of your stack — it's a bookmark. And once a quarter, look at what you're paying for and cancel anything you haven't opened.
That's it. A stack isn't a collection to admire. It's a small set of tools that each earn their place, reviewed often enough that the dead weight never piles up.
The ten-minute quarterly review
The whole system stays clean with one small ritual, done four times a year. It takes about ten minutes and it's the only maintenance a good stack needs.
Open your list of subscriptions — the actual billing page, not your memory, because memory is generous about tools you've abandoned. Go down the list and ask one question of each: have I opened this in the last month? Not "might I need it," not "is it nice to have." Have I actually opened it. If the honest answer is no, cancel it. You can always resubscribe; tools don't disappear, and the friction of resubscribing is far smaller than the slow drain of paying for things you don't touch.
Then look at what's left and check for overlap. Any two tools doing the same job? Keep the one you reach for instinctively and drop the other. Redundancy in a stack isn't safety. It's just two bills for one outcome.
That's the entire review. Ten minutes, quarterly, and your stack never quietly bloats into the thing you're embarrassed to look at. The people with clean, effective stacks aren't more disciplined than you. They just run this small check often enough that the clutter never accumulates.
If you're feeling behind because your stack is smaller than someone's screenshot online, don't. This is not a strategy. That's a tab problem, and you've simply avoided it. A short stack you use beats a long one you manage.