You can use them for practicing techniques. Have cards which say: use X technique today. You need to actually do that rather than spend 1 minute thinking about it. Which is suprisingly hard. I suspect it works much better if you have some system to guide you in generating new ideas e.g. Zettlekasten. I suspect it could be even better if the method was incorporated into the software itself. Maybe create links between cards as well, and have some repititions where you explore the graph surrounding a card?

I'm also unsure if the spaced repition timings are optimal for drilling techniques. Does anyone know the relevant literature?

For skills, I created software that reminds me to practice the skill using spaced repetition.

When the skill comes up, the program prompts me to practice  that habit in one of a couple of ways:

1. Mindfulness - I create a reminder about that habit that day, and use that to remember.
2. Visualization - I use one of the "habit installer audios" I have to go through visualization exercises of me using the habit.
3. Deliberate Practice - I create a specific period of time where I'm doing a bunch of tasks that specifically require that habit.
4. Conscious Practice - I set aside like an hour where I find any excuse to use the habit.

This allows me to gradually get better at using the habit.

For mental frameworks, I do something I call progressive expansion in Roam Research. I use the Roam Toolkit to set different models on a spaced repeitition schedule. Then, instead of testing myself when they come up, I try to expand on them in one of the 4 following ways:

1. Elaboration: How is this similar to other mental models I have?
2. Distinctiveness: How is this different from other mental models I have?
3. Application: How can I apply this mental model to things I'm doing right now? How could I have used it in the past?
4. Questions: What questions do I have about this model?

This does two things. First, it ensures I'm creating new neural connections to the model, which means it will come up more in my thinking. Second, I'm creating new Roam links to the model, which ensures it will come up more in my Roam databse.

My own suspicion in these cases is that these techniques don't work and are unsustainable hacks that temporarily shift the equilibrium of my thought. Or perhaps they do work but the mind adapts to them (like drugs) and so they'll lose their usefulness. After all, I don't forget useful things that I commonly use. even when they become implicit knowledge, I can still feel that the implicit skill is there and see its results.