[Userscript] Wanikani Item Inspector

Version 1.12.1 is now available - please download it at the top post.

This version has new information in popups.

  • Percentage Incorrect Total
  • Percentage Incorrect Meaning
  • Percentage Incorrect Reading

Exports now supports this information:

  • Meaning Mnemonics
  • Reading Mnemonics
  • Meaning Hints
  • Reading Hints
  • Context Sentences
  • Meaning Notes
  • Reading Notes
  • User Synonyms

A settings indicates how to handle context sentences:


@neversleep you were looking for this. It is here. The setting you need is Separate Rows JP First. (the default)

f you use Item Inspector for studying your leeches you should try the Leech Training script. This script will set a quiz sessions designed specifically to the study of leeches. This is a nice complement to Item Inspector that I highly recommend.

Starting from now Item Inspector can setup quizzes in a way that is close to Leech Training . This comes handy in the following circumstances.

  • Settings : Item Inspector lets you configure all aspects of the quiz session. Leech Training doesn’t have settings at all.
  • Tables : Item Inspector can setup Leech Training style quizzes for every table. For example if you want to be quizzed on Last Failed review or on JLPT N3 kanji, with Item Inspector you can. Leech Training does only leech training.
  • Quiz types: Self Study Quiz supports more quiz types: There is English to Japanese and audio quizzes. Leech Training supports only Japanese to English or Reading types.
  • Alternative : Users that are dissatisfied with Leech Training may consider Item Inspector instead. @yukinet maybe you are looking for this.

In Item Inspector the Leech Training emulation quizzes are setup with the Random Selection button. button When you click it items are randomly selected from your current table and displayed on the screen. It your settings require a Leech Training style quiz, Random Selection will set it up for you. Then you may use one of the Self Study Quiz buttons to be quizzed on these items.

To undo a random selection click the button again. The full table is displayed back.

You get a new selection every time you use the Random Selection button. You may have one new quiz every day, week or whatever your schedule is.

Leech Training has a concept of “clearing the leeches”. It requires you to get the leech (and visually similar items) three times to clear the leech. Self Study Quiz has no such concept but it will give you the option to be quizzed on the same items as many times as you want. The decision of whether you had the items right often enough is yours.

You may configure the Random Selection with these four settings. You may also define your own settings on what you consider a leech. In Leech Training the definition of leeches is built-in and can’t be changed.


  • Random Selection : This setting sets the size of the random selection. The default 0 value means to select as many items as is needed to fill the screen.
  • Only Once Before Next Review : If checked this setting will select an item only once since the last review. Leech Training does this better because it can track reviews and quizzes done on several devices. Item Inspector will track your activity on a per browser and device basis. There is no hope to improve on this for technical reasons.
  • Add Similar Items : If checked this setting will cause up to three visually similar items to be added with each randomly selected item. Leech Training does this. I can’t compare both scripts because I don’t know how Leech Training does its visually similar item selection.
  • Visual Similarity Threshold : This settings is used primarily to select which Lars Yencken visually similar kanji are displayed in popups. Leech Trainer emulation uses Lars Yencken visually similar data and honors this setting when adding visually similar items.

Item Inspector defines “visually similar” as follows.

Common Rules

  • If Only Once Before Next Review option is selected the visually similar item can be selected only once befor it is reviewed.
  • Items must have had a lesson to be selected. You won’t be quizzed on an item you have not learned.
  • To be similar two items must be of the same type (Radical, Kanji, Vocabulary)


  • There no visually similar items for radicals


  • Visually similar kanji according to Wanikani and Lars Yencken are eligible to be selected.
  • The LY similarity threshold is honored.


  • A kanji is considered similar to itself in vocabulary context.
  • Visually similar kanji according to Wanikani and Lars Yencken are eligible in vocabulary context.
  • The LY similarity threshold is honored.
  • A vocabylary item is similar to another if all kanji are similar to each other.
  • The similar kanji need not be in the same order.
  • The number of kanji must match.
  • The similarity correspondence must be one-to-one. (You can’t have multiple kanji.similar to the same kanji)
  • The similar kanji need not be in the same order.
  • Non kanji characters are ignored.

To sum it up, Leech Training does fewer things but it excels at what it does. If you just need to be trained on your leeches this is likely your best option. Item Inspector has more settings but it is handicapped by technical limitations. There is no multiple device support and no concept of “clearing the leeches”. There is no hope to ever include these features. Use Item Inspector when your needs go beyond what Leech Training can do.