I’ve implemented a screensaver of all of your leeches. You can see it by visiting https://wanikanitools-golang.curiousattemptbunny.com.
Animated gif (be patient):
To set this as your screensaver try - let me know if you have found a tool that lets you use this page as a screen saver!
that’s a really cool idea and all this coding stuff perplexes me. you’re all flippin’ witches, I swear. 
you’re great! <3
Great idea! I’ll be making my own wallpapers for my iPhone if things come to the worst.
Of my first three items shown, two of them I’ve burned. And I had a perfect 8 for 8 record on both of them. So something seems wrong with your leech algorithm.
Hmm. Not sure what’s up with that, @seanblue. I’ll take closer looks at the data in a bit. I might need access to your API key to debug it. We’ll see. Thanks for giving it a try!
@seanblue I’ve tightened up the logic for the leach scoring. Try again. It’s still possible for burned subjects to show up because so far I’m not retrieving the level data.
I’ve added some fancy CSS transitions to the page, so things fade in and out. I tweaked the timing so that information is revealed slowly enough for my to guess at the reading and meaning.
Here’s a still to give an idea as to what it looks like now:
Note that there’s no loading-animation yet, so if you’ve lots of active subjects then it might take a little longer.
Here’s the leaches filtering and scoring logic that I’m using. I’m open to suggestions for improvement:
Oh. I think there’s a bug whereby only my API key is used. I’ll fix that in a bit.
It still only works for my api key. I’ll fix that later.
Right now I’ve improved the display a bunch. It now displays the reading as furigana! 
and, well, I’m kind of cheating with the furigana so sometimes the kanji <-> furigana partitioning is approximate:
@seanblue Ok. I’ve made it so that the api_key works. Try again and let me know how it goes.
@hitechbunny, this is brilliant! Thank you for sharing it. I installed the Mac OSX screensaver software and now it flashes leeches whilst I’m doing other things at my desk.
What is the time interval between displaying the kanji, meaning and reading? it seems to be quite slow on my computer. Are you deliberately using those pauses to make us think about the answer?
Good job.
@wunderbunny Yes. I timed it to allow myself enough time to guess correctly for one I was struggling with. Although, I’m no longer struggling with any that show up for me on my screen saver. 
I wonder, is this cheating to practice outside of the SRS intervals??
No, it is not cheating to practice outside of the SRS intervals. It’s called studying. I put my leeches into an Anki deck and review them at shorter intervals (1 min, 1 min, 10 min, 4H, 8, 1 day, 2 days, 3 days, 4 days, 7 days).
I found that the one week and two week intervals for the guru-ed items were so long that I forgot them. With my Anki deck, I add any missed item that does not move into guru, master or enlightened.
However, it is cheating if you are revising your items less than one hour before they are to be reviewed in WaniKani. You should use the WaniKani as a diagnostic. The review is answering the question “Do we know it yet?” It is not a high stakes exam. There is nothing shameful about getting it wrong.
How do you go about making sure what you’re reviewing in Anki isn’t “just” before WaniKani brings it back again? Doing them at different times a day?
@chezmax, I usually go through the Anki deck after I completed my morning review. It is a matter of luck if I encounter an item which is subsequently reviewed in the next hour or on the same day. It is a rare occurrence.
With my current Anki review timings:
If a guru 1 week item fails, it drops back to the apprentice one day review. I add it to Anki and review it at 1min, 1min, 10 min, 4 hour, 8 hour, 24 hour. When WaniKani next presents that item (one week later), I would have seen the Anki 8 hour review. Anki then continues with 24 hour, 2 day, 3 day, 4 day, 7 day reviews. When WaniKani presents it for the two week review, I would have seen it four more times.
Hmm, at first I got a word that I would qualify as a leech, but then the following two were two kanji from the level that I am currently on.
I’ve only got them to apprentice level 2 at present, but I haven’t answered them incorrectly a single time.
How are leeches qualified? Surely both really new items and items with 100% correct answers should be excluded?
The idea is cool though!
I was wondering about that too… what are you qualifying as leeches in your script?
And since you can’t use the website as a screensaver on Windows 10, I wonder if anyone knows of a way to do it on Windows 7. (I’ve never heard of using a website as a screensaver before! That sounds like magic to me.)
Cool idea.
Yeah. I just guessed at a way of scoring subjects and then took the worst 100: wanikanitools/app.jsx at 4fbfe484fb3cbdeee6844bb17affa5388f189994 · curious-attempt-bunny/wanikanitools · GitHub
I think instead I’ll copy the scoring used by Leech Detector and use all subjects scoring worst than a certain threshold.
I’ve also been meaning to add a way of listing everything the screensaver will show. It’s all a matter of making the time to work on it.
I noticed that loading the screen saver was taking 20-60 seconds for some folks. I hadn’t noticed this slow down for myself because I’m at level 5. Anyhow, I’ve speeded things up by making proper use of the v2 api and only pulling down what’s changed. Loading should be sub-second now for folks that are re-visiting the site. Although right now every time I deploy an update I’m destroying the cache … um, so the second visit and onwards should be fast for folks. Thanks for your patience!
Thanks for using your time to make something for others 
