or upload a text file:
or load from Project Gutenberg:
Tokenize all the prose, tag each word against four open lexicons (NRC emotion, Brysbaert concreteness, AFINN sentiment, SUBTLEX baseline frequency). Source can be a Bluesky archive (CAR fetch from the user's PDS) or an uploaded plain-text / markdown document. All deterministic; no inference; nothing leaves your browser.
Resolving handle…