Why I Built Burly

I have spent most of my career building software, and the projects that stick with me are almost always the ones that started as a personal annoyance I refused to keep tolerating.

Burly is exactly that.

It did not come from a market analysis or a roadmap. It came from the small, daily friction of clicking a work link from Slack and watching it open in my personal browser, logged into the wrong account, next to the wrong tabs.

The annoyance was tiny and constant

No single instance of this problem is a big deal. A link opens in the wrong place, you notice, you drag the tab over or copy the URL and paste it into the right window.

Ten seconds, maybe.

But you do it dozens of times a day, and the cost is not really the ten seconds. It is the interruption. You were reading something, and now you are doing tab logistics instead. The contexts you worked so hard to keep separate collapse into each other anyway.

I kept waiting for the friction to feel small enough to ignore. It never did. It just quietly taxed every link I clicked.

The tool I wanted did not quite exist

There are apps that let you set rules: send anything from this domain to that browser. I tried that approach and it never fit how I actually work.

My routing is not really about the domain. The same link might belong in my work profile in one moment and my personal profile in another, depending on what I am doing. A rule engine forces me to predict that in advance, and to maintain the rules forever as my life changes.

What I wanted was simpler and more honest: at the instant a link arrives, just ask me where it should go, and make answering instant.

So that is what Burly does.

macOS does have pieces of this. It lets youset a default browser, and Safari documentsprofiles with their own external-link behaviour. What neither gives you is one picker that routes across every browser you run.

How does Burly compare to the other macOS link routers?

I did not build Burly in a vacuum, and it is not the only app in this category. The honest summary is that every one of these tools routes links well; they differ on what they can treat as a destination. Here is where each one actually lands, and the detailed breakdowns live on the Burly comparison pages.

Burly vs. Browserosaurus

Browserosaurus is the app most people in this category started with, and it is the one Burly owes the most to: click a link, get a grid of browsers, pick one with a keystroke. Its author retired the project in 2025, and the repository is archived. It also chose between browsers rather than between profiles, so it could not tell a work Chrome from a personal Chrome, and it shipped on Electron rather than native code. Burly is native, maintained, and profile-aware.

Burly vs. Choosy

Choosy has the deepest rules engine of the group at $10, and it routes more than links: the Share menu, Handoff, and AirDrop all go through it. If you want to declare in advance where every domain belongs, Choosy is the better fit. It reaches Chromium and Firefox profiles but cannot target a Safari profile, which is the exact case that made me build Burly.

Burly vs. Velja

Velja is $8, beautifully made, and built around per-host and per-app rules. Same shape of trade-off as Choosy: excellent if you want a rule engine, and it reaches Chromium and Firefox profiles, but a Safari profile is out of range.

Burly vs. Bumpr

Bumpr is the cheapest at $2.99 and the most focused: it picks between browsers and between mail clients. That is a genuinely different job, and it does it well. It does not target profiles at all.

So the split is not really about quality. It is about whether the profile is the destination. If you keep separate Chrome profiles, most of these will serve you. If your work life lives inside a Safari profile, the field gets much shorter. OpenIn is one of the few that can open links in a Safari profile, and it is worth a look. That gap is what Burly was built for.

How it works

Burly becomes the system handler for the links you click. When you open one, a fast radial picker appears and you choose the destination with the keyboard: Safari Work, Chrome Dev, whatever profiles you keep.

It discovers your browsers and profiles automatically and keeps that list live, so I never have to configure anything by hand. It works across Safari, Chrome, Brave, Edge, Vivaldi, Arc, and the Firefox family.

And because being asked every single time would become its own annoyance, there is a modifier key that skips the picker and reopens your last-used destination. Most of my links follow a streak: several to work, then several to personal. The bypass makes the common case free and saves the choice for when it actually matters.

Keyboard-first, because speed is the whole point

A router that routes links slowly is worse than no router at all. If the picker took even a beat too long to appear, or if it forced me to reach for the mouse, I would stop using it and go back to dragging tabs.

So the picker shows up the instant a link arrives, and every choice is a keystroke away. That constraint drove more of the design than any feature did. The bar was simple: Burly had to be faster than the habit it was replacing, or it did not deserve to exist.

Private by default, like everything I ship

A tool that sees every link you click has to be trustworthy, so Burly keeps everything on the Mac. There is no account, no cloud, and no analytics, and no URL ever leaves the device. It is a native menu bar app, signed and notarized.

This is the same instinct behind the other small tools I make. The links you open are some of the most revealing data about your day. The right move is not to protect that data carefully. It is to never collect it in the first place. Data you never gather is data you can never leak.

Is a problem this small worth solving?

Why a problem this small was worth solving. It would be easy to dismiss Burly as trivial. It routes links. That is the entire pitch.

But I have come to believe the small, constant frictions are exactly the ones worth removing, precisely because they are constant. A big problem you hit once a month gets your full attention and a real workaround. A tiny problem you hit forty times a day never rises to the level where you stop to fix it. It just wears a groove into your attention.

AI has made it cheap enough to build a small native app that fixing these little grooves is finally worth an afternoon. That changes the math on what deserves to exist. A tool does not have to be ambitious to be worth building. It just has to remove something you were tired of tolerating.

The test I keep applying

Burly passed the only test I really trust for a personal tool: I reached for it the next day without thinking, and the friction it replaced was simply gone.

I built it for myself, and it turned out other people run their browsers the same anxious, context-separated way I do. That is usually how it goes. Scratch a specific enough itch and you find out you were never the only one scratching it.

Every link, in the right browser profile. It is a small promise, and keeping it well is the whole product.

← All postsNext: Small, Local Software →