A roundtrip failure test may be good - add shortTag to either the encrypt or decrypt, but not both.

Truthiness tests may be good. What's the expected behavior if these flags are "true"? "false"?

Generally speaking, we should remove jsdocs from ts files and replace them with typed params

params are typed - rm jsocs (except description)

An alternative:
`
// Define supported languages with type safety
const languages = ['en', 'es', 'ja', 'zh', 'fr', 'it'] as const;
type Language = typeof languages[number]; // Creates type literal: 'en' | 'es' | 'ja' | 'zh' | 'fr' | 'it'

const wordsForLang: Record<Language, Array> = {
en: Mnemonic.Words.ENGLISH,
es: Mnemonic.Words.SPANISH,
ja: Mnemonic.Words.JAPANESE,
zh: Mnemonic.Words.CHINESE,
fr: Mnemonic.Words.FRENCH,
it: Mnemonic.Words.ITALIAN,
};
`

Then in your KeyOptions interface, language?: Language

See above. If you don't want to make that refactor, consider uniformity of string literal definitions - you define the KeyAlgorithm string literal above with keyof typeof, but Language (analog) is defined on the itnerface directly

rm jsdocs

The optional chaining after toUpperCase is extraneous

extraneous optional chaining after toUpperCase

extraneous optional chaining after upperCase()

extraneous optional chaining after upperCase()

extraneous optional chaining after upperCase()

This could lead to type errors since it's initialized as null.

Should we be ensuring that this.credentials is a Credentials, and not some other 'CredT'?

Is this one of the subject to change parts? xPubKey unused (use commented below)

unresolved

leaving this in here?

Please add a clarifying comment above this conditional

This makes it seem like !this.#keygen≥isKeyChainReady() is the primary flow. If that's not the case, consider:
if (!this.#kygen.isKeyChainReady()) { return null };
... primary flow & return key

n is number of participants, and m is number of required signers - right? Brief inline documentation in this file would be good, especially for ambiguous fields

Same comment in TssKeyGen class below

If you use the default value of extra (or if it's undefined), then you end up with a ':' at the end of your string.

const myParams = {
	id: '123',
  partyId: '456',
  m: 3,
  n: 5
};

const myStr = [myParams.id, myParams.partyId, myParams.m, myParams.n, myParams.extra].join(':');
const myStr2 = [myParams.id, myParams.partyId, myParams.m, myParams.n, ''].join(':');
console.log(myStr);
console.log(myStr2);

If that's not the intended behavior -
Suggestion:
const myJoinCodeArr = [this.id, partyId, this.m, this.n];
if (params.extra) { myJoinCodeArr.push(params.extra };
const code = myJoinCodeArr.join(':');

If opts.encoding, then is it ensured to be in your list of provided values? Would validation be useful here?

const encodingSupported = ['hex', 'base64', 'utf8', 'binary'].includes(opts?.encoding);
const encoding = encodingSupported ? opts.encoding : 'hex';

Alternatively, as an args check?

see note in createJoinCode comment above

What should happen if password isn't provided in opts, and there's no 'more'?

It will be an empty string on the reassign - is that alright?

The string format would be nice: 'id:partyId:m:n:keygenSession'
And in the restoreSession below it.

unresolved

A comment would be really useful here