Kirby CMS uses plain .txt files for content. Since the bare .txt files should not be accessible with a browser, one normally uses a path matcher in Caddy and then denies requests based on a path, like so...

path *.txt /content/* /site/* /kirby/* /.*

But what if I want a /robots.txt file? Turned out to be a simple answer, but it took me a while to find it. I'm writing it down here in case anyone else might need it. Here's the whole block from my site's Caddyfile:

(kirby) {
    php_fastcgi unix//run/php/php8.3-fpm.sock
    @blocked {
        path *.txt *.md /content/* /site/* /kirby/* /.*
        not path /robots.txt   # <----- Here's the important part
    }
    respond @blocked "Not Found" 404 {
        close
    }
}

With that one extra line, when you add an import kirby line to a server block, /robots.txt will remain accessible.