I have referred this entry and its comments to Kim.

Yes, hiding rare irregular kanji forms is our usual practice. If 含羞む were in other refs, we'd tag it as rK.
Jim, Brian's suggestion to contact the people at jisho.org and alert them to the uk/sK issue seems like a good idea. At the same time, you could remind them about recently added field tags (e.g. ski, film, tv, rommyth) which still aren't supported on jisho.org.

Since the kanji form is in プログレッシブ it should be searchable. it is very rare and I agree it's probably best hidden. [uk] should stay as that has some impacts on apps using this data for parsing.

In a fork, Brian wrote:

  Re: Robin
  It is my understanding from reading here that the jisho.org maintainers are pretty responsive.  Is there a chance someone with existing connections could make a request:

      Please don't display the "usually kanji" or other kanji-dependent tags (eg. [gikun]) when no kanji are displayed due to [sK]?

  Re:Marcus

  I had/have the exact same viewpoint. From the vantage of studying kanji, rather than studying vocabulary, I like seeing these forms even if they are rare.  But I defer.
  Here's the github discussion for reference:
  https://github.com/JMdictProject/JMdictIssues/issues/75#issuecomment-1435712638

We decided a while ago that we didn't want both [iK] and [rK] tags on the same form, so the alternatives were to tag it [sK] or to drop it altogether.

We've been dropping rare and irregular forms which don't produce any n-gram counts (see: Robin's comment on 1238790). Since 含羞む gets a few hundred counts, I don't think we should dismiss the idea that people will search for it. I think having these search keys in the database can also be useful for programs that are designed to parse and tokenize text.

Many entries are affected by this policy. See for example comments on entry 2854734