I’d not bought a new app for ages, and since there were a bunch of discounts last weekend, I decided to raid the piggybank and buy BTT – mainly because it has been mentioned here a few times and because it was going cheap.
But what to do with it?
I quickly found @nontroppo’s one-click compile, which will come in handy. Thanks for that. It was also instructive for a BTT noob, and introduced me to the Hyper Key – at last a use for Caps Lock \o/
Using ideas from that, I was able to construct my daily Scrivener startup, which involves opening a few projects, applying templates, and one or two other things. It’s quite magical to sit back and watch it all unfold with a single keystroke.
Another nice touch – though not Scrivener-specific – was being able to emulate a mouse middle-click on the trackpad.
Anyway, I realise I’ve only scratched the surface of what BTT can do, and it’ll take a while to go through the docs, so I thought I’d ask for suggestions from fellow users here.
So I select an image in the binder, then gesture down+left to trigger the menu to replace that media files (I store my figures in the Binder, and use this when I update them externally). I have gestures to run my Build automation etc.
Key sequences to trigger Scrivener styles: BTT can run actions on sequences of key presses, and I use this to activate styles: – so ``` (which is markdown to start a code block) triggers my block code style automagically. I mostly do this for paragraph styles.
Run an AI editor: I use LM Studio to run a local LLM, and use BTT to take my text selection in Scrivener and run it with a custom system prompt through the LLM and paste the text back in Scrivener.
Extend keyboard bindings: many apps can do this (and macOS can do it natively) but I just centralise everything in BTT. So for example I use BTT to rebind ⌘b to use bold style.
There are so many features that I still get lost in BTTs action menu but there is enough simpler stuff that I can’t live without it…
Gestures: Ditto. The reason I still use Firefox is because of gesture support. I’d like to move to Safari – simply because it’s one less app, and I’ve reached the stage of life where simplicity overrides most other things – and have already setup a few to try with Safari. I’m going to have to monitor my behaviour for a while to discover where else I can use gestures.
Key sequences: Ah, that’s a good idea. For example, I use Scrivener annotations with a TODO: prefix. This will speed that up and avoid typos.
Floating menus: I looked at these briefly during my initial exploration. I’m pretty good with Scrivener’s shortcuts – the ones I need, obvs. – but there are a few functions I use rarely that I can see this being handy for. Also for some of the GUI actions that then need mouse action, like Containers, for example.
LLM back and forth: This would certainly make an LLM more useful when writing. I tried LM Studio after you mentioned it before, but soon lost interest. This sort of ‘integration’ is probably more viable.
Keyboard re-bindings: Yes, I think it makes sense to centralise those. I only have a dozen or so atm. This is similar to the way I moved text replacement into espanso.
You can do just anything with BTT. @nontroppo 's useful list could be continued endlessly. I’m not going to do that, even if it tempts me.
Instead, I’ll tell you this: I have set up a “reverse keyboard concept” with BTT (originally only for Scrivener). This means that all letters are typed when the key is released (not when the key is pressed as usual).
This enables the following: If a letter key is tapped briefly (as in normal writing), the letter is typed. If the same key is held down a little longer, a different action is triggered. Simple example: Holding “I” a little longer shows or hides the inspector.
I (almost) no longer need modifiers. This is wonderful for me because I can keep my eyes on the screen instead of constantly looking at the keyboard. It promotes my writing flow.
Ha that is both totally insane and eminently logical at the same time
I use 3 assistants, a scientific editor to improve clarity, an assistant to answer questions, and an editor to shorten text, they all work on selected text in any app and use whichever model is loaded in LM Studio (server port 4891, an oblique Orwell reference!).
I also want to mention if you use Alfred, a brlliant workflow called Ayai which also has a nice set of actions that works with LM Studio (or any other LLM):
I switch between 3 layouts (mini, midi and max with the dice icons), two links icons “import” either files or aliases to the Research folder, and links to project search, snaphot with title and quick search: scrivener-menu.bttpreset.zip (92.3 KB)
– so 
but there is enough simpler stuff that I can’t live without it…
