From beta to launch: building MOLTamp in public

MOLTamp has been in private beta for several months. This post is a midpoint reflection: what we have learned, what is changing as we get closer to public launch, and the philosophy behind why we are building it the way we are.

If you are a beta user, this is the rough state of the union. If you are not, this is an honest look behind the curtain.

What private beta actually means

"Private beta" makes it sound more exclusive than it is. The reality is that we wanted real users running the app in real conditions before we put it on Product Hunt and got slammed. The beta was a way to find the bugs that only show up when the user has 200 widgets and a custom skin and a weird shell config and runs Claude Code with three subagents simultaneously. That stuff is impossible to find in synthetic testing.

The beta has been small — under 200 active users. Small enough that I can actually read every Discord message. Big enough to surface the bugs that matter. The tradeoff between "more users finds more bugs" and "more users means more support burden than I can handle alone" landed in this range almost by accident.

What we have learned

A few specific things that surprised me:

People skin much harder than I expected

I built MOLTamp expecting most users to pick one of the bundled skins and stop. Wrong. Most active beta users have customized at least one variable, and a meaningful fraction have built their own skin from scratch. The skinning system is the most-used customization surface by a wide margin.

This changed how I think about the priority order for features. The skin browser, color customization, and skin export pipeline got way more attention than I originally planned because that is where users actually live.

Vibes matter more than features

The beta users who use MOLTamp every day are mostly people who tell me "it makes me want to sit down at my desk." That is a weird thing to optimize for as a product person — it is not measurable, it is not in any KPI dashboard, it is not a bullet point on the homepage. But it is the thing that distinguishes "people who installed it once and forgot" from "people who use it daily."

I have leaned into this. Every feature decision now goes through a "does this make the room more inviting" filter, not just a "does this save keystrokes" filter. Sometimes those are the same. Often they are not, and the vibes filter wins more than I expected.

The community gallery is the killer feature

I built the community gallery as a "version 2 thing" and ended up shipping it in version 1 because beta users kept asking. It is now the thing I get the most positive feedback about. Not the skin engine itself. Not the widget system. The act of being able to browse what other people have built and install their work in two clicks.

This makes sense in retrospect — the gallery is where the network effect lives. A skinnable shell with no community is just a customizable terminal. A skinnable shell with a thriving gallery is an ecosystem.

What is changing for launch

A few things will be different when MOLTamp goes public.

Pricing

The model is staying the same: free forever, $20 one-time to remove the periodic nag. No subscription, no tiered pricing, no "Pro" features locked behind the paywall. Every feature works in both versions. The $20 just removes the popup.

The reasons for this pricing:

I hate subscription fatigue and refuse to add to it
The product is small and focused enough that there is no "Pro tier" of features that makes sense
The nag-removal is a fair value exchange: if you use it enough to be annoyed by the nag, $20 is nothing
Free forever lowers the barrier to trying it, which matters for vibe-driven products

Refund policy

30 days, no questions asked. We are confident in the product and confident enough that someone unhappy at day 25 should get their money back. I would rather refund a few people than have anyone feel trapped.

Documentation

The documentation has been "ask in Discord" through the beta. That works for 200 users. It will not work at scale. We are writing real docs — the skinning guide, this blog, the upcoming widget tutorial series. Goal is "you can find the answer to any common question without asking a human."

Support model

Discord stays the primary support channel. Email for things that need it. No support tickets, no help desk software, nothing that adds bureaucracy between you and an answer. As long as the user count stays small enough for me to handle this personally, that is the model.

What is NOT changing

The free tier with the nag — staying
The core feature set — staying
The opinionated visual identity — staying
The "vibes are the feature" framing — staying

If anyone tells me I should add a Slack integration to compete with X or build a "team plan" because every other dev tool has one — they are misreading what MOLTamp is. It is a single-developer aesthetic-first tool. That stays.

What I have learned about building in public

I have been semi-publicly building MOLTamp on Twitter and Discord for months. Some honest observations:

Showing screenshots gets traction. Showing features does not. A screenshot of a beautiful skin gets 100 likes. A screenshot of a new widget gets 5. Visual is the universal language for a visual product.
The same audience asks for "stop showing off, ship the Windows build" and "post more screenshots." They are different sub-segments and they are both right.
Public roadmaps create pressure that is sometimes useful and sometimes unhealthy. I used to maintain a public roadmap with dates. I stopped because the dates created pressure to ship things half-done. The current roadmap is intentionally vague.
The community is much more patient than I assumed. I was scared to ship a beta with rough edges. The beta users have been kind about rough edges and direct about real problems. Nobody has been mean. This is uniquely valuable and I did not expect it.

What is next

The next 60 days are the run up to public launch. Specifically:

Windows builds shipped (about 3 weeks out)
Linux AppImage shipped (a week after that)
Documentation site polish
Product Hunt launch prep
The blog you are reading right now keeps growing

If you want to be part of the launch ramp, the Discord is where it happens. Or just keep using the beta and tell me when something annoys you. Both help.

Thanks for being here for the messy part.