Extension to Use Keicho on Scrapbox

• A tool that allows you to use Keicho in Scrapbox.
• /takker/PopupMenu to ask Keicho about the selected range by (takker)
  • You can use it with the UserScript and tamperMonkey askKeicho.
  • (The implementation is fast and impressive (blu3mo))
  • As usual, I was poking around with the development tools and found an API, so I tried making it and it worked (takker)
    • I’m hitting the API without permission, but I hope it’s okay.
      • (If I write it here, (nishio) should respond)
      • Using someone else’s project as a bulletin board might not be a good idea.
        • (It’s totally fine (blu3mo))
        • (Thank you (takker))
        • (nishio) What an interesting turn of events.
      • (nishio) First and foremost, I find it very interesting that Keicho can be used in Scrapbox.
        • On the other hand, I didn’t anticipate this usage at all, so I’m curious to see what you’ll do next.
        • Well, even if something really strange happens, it should only result in specific conversations being inconsistent and shouldn’t affect other users, so I think it’s fine.
          • I skimmed through it, and it seems to be doing the same thing as the web client, so it should be fine.
          • Why do you need a browser extension? It would be more convenient if it could be done without it.
            • When trying to access keicho.herokuapp.com from scrapbox.io, it gets blocked by CSP (takker)
              • It’s a security measure to prevent unauthorized data transmission to malicious external servers.
            • To bypass that, I’m accessing keicho.herokuapp.com via tampermonkey.
            • cf. /ci7lus/たのしいScrapboxUserScript#5f63841bae0f140000027cd6
            • (nishio)
              • Since you can access keicho.herokuapp.com from keicho.netlify.com, you should be able to perform cross-origin access, right?
              • See /nishio/Flask-CORS
              • The server side has already been taken care of, so if you fetch in CORS mode on the client side, you should be able to access it from Scrapbox.
            • It’s not possible because scrapbox.io, i.e., the client side, has restrictions (takker)
              • @Chrome
                • 
              • By the way, I was able to access it for some reason in Firefox…
              • (nishio) Ah, I see, so Scrapbox restricts the communication destination with a whitelist.
        • In some cases, a dedicated API may need to be created.

keicho.json

{"talk": "PJAcJJ3AKxo407IhZDgm"}

Usage test

• I want to think about how Bots can live in Scrapbox.

• (/nishio/nisbot) What kind of “Scrapbox” is that “Scrapbox”?

• There are both synchronous and asynchronous conversations in the discussion space.

• (/nishio/nisbot) What kind of “conversation” is that “conversation”?

• It’s like stacking building blocks of thoughts through the exchange of information.

• (/nishio/nisbot) What kind of “building blocks” are those “building blocks”?

• Thoughts

  • It would be nice if Keicho could perform a depth-first search and express it as nested bullet points.
    • I wonder if it has that information internally?
      • (nishio) It might be possible because it has a mapping of keywords to the input when they appear for quoting past sentences.
  • Applying Keicho to random text in Scrapbox might be surprisingly good.
    • It seems like a good way to pass the time.
    • From there, you can start a conversation when you feel like it.
    • It would be great if ScrapBot could do this kind of thing.