I have a one-liner similar to Marc’s but with a simpler Regular Expression and ~20% faster according my benchmark (Chrome 89).
const kebabize = (str) => str.replace(/[A-Z]+(?![a-z])|[A-Z]/g, ($, ofs) => (ofs ? "-" : "") + $.toLowerCase())
const words = ['StackOverflow', 'camelCase', 'alllowercase', 'ALLCAPITALLETTERS', 'CustomXMLParser', 'APIFinder', 'JSONResponseData', 'Person20Address', 'UserAPI20Endpoint'];
console.log(words.map(kebabize));[A-Z]+(?![a-z]) matches any consecutive capital letters, excluding any capitals followed by a lowercase (signifying the next word). Adding |[A-Z] then includes any single capital letters. It must be after the consecutive capital expression, otherwise the expression will match all capital letters individually and never match consecutives.
String.prototype.replace can take a replacer function. Here, it returns the lowercased matched capital(s) for each word, after prefixing a hyphen when the match offset is truthy (not zero – not the first character of the string).
I suspect Marc’s solution is less performant than mine because by using replace to insert hyphens and lowercasing the whole string afterwards, it must iterate over the string more than once, and its expression also has more complex look aheads/behind constructs.
Benchmark