I'm a really big fan of Alfred. Hitting a hotkey to summon a text input, multi-function app launcher is a paradigm I fell in love with when I first found it. I have an old entry from 2010 covering how Quicksilver was the killer app for me that took working on a Mac from good to great. At this point, I use it even to switch apps so that I don't have to mentally track or read the list of recently used apps. Rather than ⌘+tab, I summon Alfred and type ch to get to Chrome, v to get to VS Code, or t to switch to terminal.

I know the cool kids have switched to Raycast, but I don't trust that it's a sustainable business. I've seen them change the paid features several times. Initially it was just "We'll figure it out later, but we promise free forever". Then they moved to only charging for team based features. The current version charges for sync, AI, and a few other features - but still has a free tier that includes everything useful. None of these look anything like a workable funding model to me given the size of the company: it's 40 people and VC backed. There's a bait and switch coming. For something that's not core to my workflow, I'm sometimes ok with that. Canva's new freemium Affinity, for example, is something I use but don't really expect to last. Alfred, though, is right at the center of how I use my computer. For a program that important, I want either a healthy looking open source project or to pay for it in a way I think is sustainable. Running with Crayons, the company that sells Alfred, is two or three people. They charge for major versions every few years. They're sustainable in a way that there is a way that Raycast clearly is not. Even better, they're using a model where I don't have to pay every month.

Anyway, I decided to spend a bit of time polishing workflows. Quite a few of these overlap with something that already exists. Those existing solutions, though, either didn't quite work the way I wanted or for other reasons didn't work for me. For example, php used to be pre-installed on macs and some old workflows rely on it. Now I'd have to install it, and I'd rather not.

alfred-password-gen is a memorable password generator. This is the only one of my workflows that has unique functionality. Macs used to have this built in, but it disappeared quite a while back and I made this as a replacement. Random strings are a great choice for a lot of passwords. When you have to type something on a TV or a printer, though, a memorable password can be quite useful. SENTRY&14Allocations=4 is an example of the sort of password it generates.

alfred-uuid returns a v4 uuid. uuidgen | pbcopy gets me the same thing, but I often want a uuid in unit tests or other dev workflows, and it's nice to have it right in Alfred. A few of these exist, but they have external requirements like Node. It was easy to recreate without dependencies.

alfred-unicode searches through unicode. It uses the Unicode CLDR Project projects dataset, so things like 🎉 will match any of "awesome", "birthday", "celebrate", "celebration", "excited", "hooray", "party", "popper", "tada", or "woohoo". There are several workflows like this, but none of them worked quite the way I wanted.

alfred-timestamp converts unix timestamps into human readable values, both local and UTC. 1771081036, for example, will return 2026-02-14 06:57:16 am PST and the UTC equivalent. This isn't unique either; there are a few similar workflows out there.

alfred-encode-decode encodes to or decodes from base 64, json, url encoding, and html entities. A couple workflows like this exist, but one of my common needs is json and none of them included that.