The thing I used to do before Wand
You know the move. You're reading a dense error log, or a weird config option, or a paragraph of legal text in some API doc, and you don't fully get it. So you highlight it, Cmd+C, open a new browser tab, paste it into ChatGPT, wait for the page to load, type "what does this mean", wait for the response, read it, then try to remember what you were doing in the first place.
That whole loop takes maybe 45 seconds. But it happens twenty times a day, and the context switching is what actually kills you. By the time you're back in your original window, you've lost the thread.
Wand is the fix for this. It's a LacPointer feature, and it's one of the quieter ones — but I use it more than almost anything else.
How it actually works
Press Caps Lock anywhere on your screen. Your cursor turns into a crosshair. Hover over whatever you want explained — a block of text, a UI element, an error message, a chart label, a line of code in your editor — and click it. LacPointer reads what's under your cursor and gives you an AI explanation right there in the floating bar.
No copying. No tab switching. No extra typing. You point, you click, you get an answer.
It works across everything on screen. I've used it on:
- Stack traces in my terminal
- Unfamiliar flags in a man page
- Dense sentences in GitHub PR descriptions
- Jargon in a Slack thread I was pulled into
- Foreign language text that showed up in a dataset
- Cryptic values in an environment config I inherited
It reads context well. If you point at a specific error code, it doesn't just define the code — it explains what likely caused it and what to look at next. That's the part that actually saves time.
Copy Mode is the passive version
If you'd rather not press Caps Lock every time, there's Copy Mode. Turn it on and LacPointer watches your clipboard. Every time you copy something — Cmd+C anywhere — it automatically explains the copied text in the bar without you doing anything else.
I keep Copy Mode on when I'm doing research or reading docs. You're copying things anyway as you read, so the explanations just appear as a side effect of your normal behavior. It feels a bit like having someone sitting next to you who you can casually read things out to and they just... tell you what it means.
The one time I turn Copy Mode off is when I'm actively coding and copying snippets around, because you don't want an explanation triggered every time you duplicate a variable name. It's easy to toggle, so it's not a big deal either way.
Pushing results to Notion
This is the part that turns Wand from a lookup tool into a documentation tool. Once you've got an explanation in the bar, you can push it directly to Notion. One button. It lands in your workspace as a note.
I use this when I'm onboarding onto an unfamiliar codebase. I'll Wand things I don't recognize, and anything that seems important enough to remember, I push to Notion. By the end of a session I've got a set of notes that actually reflect what I was confused about — not a generic "getting started" guide that someone wrote two years ago.
You can also share results with teammates directly. If you Wand something and the explanation is useful, you don't have to re-explain it to someone else. Just share the result.
The Persona angle
Wand uses whatever Persona you have active in LacPointer. This matters more than it sounds.
If you're using the Coding persona, explanations come back in engineering terms — direct, no fluff, assumes you know what a pointer is. If you switch to the default persona, explanations are a bit more conversational. If you're reading something technical alongside someone non-technical and you share the screen, you could flip to a friendlier persona and get an explanation that doesn't assume background knowledge.
The Personas feature was built so that the active persona's system prompt bakes into every interaction, including Wand. So you're not getting a generic explanation each time — you're getting one tuned to how you've configured the AI to talk to you.
A quick workflow worth stealing
Here's the one I've settled on for reviewing unfamiliar pull requests:
- Open the PR on GitHub in the browser
- Switch LacPointer to the Coding persona
- Turn on Copy Mode
- Read through the diff — copy any lines you want explained
- Explanations appear in the bar automatically
- Push anything worth keeping to Notion with one click
By the time I've finished reading the PR, I've also got a mini breakdown of the parts I didn't immediately understand. It makes leaving a meaningful review a lot easier because you're not faking your way through the parts you glossed over.
One thing to know about Caps Lock
A lot of keyboards treat Caps Lock as a toggle that changes your typing behavior system-wide. LacPointer intercepts it for Wand activation, so it won't actually toggle Caps Lock on your keyboard when you use the shortcut — it captures the keypress before it gets to the OS. If you do need actual Caps Lock for something, you can hold Shift to capitalize instead. Just worth knowing upfront so you're not confused the first time you use it.
If you haven't tried Wand yet, just press Caps Lock on the next thing you read that you don't fully understand. It takes about two seconds and you'll immediately see why it becomes a reflex.
LacPointer is available for macOS and Windows — grab it at lacai.io/download. The Wand feature is available on both the free and Pro plans.