$ styleprofile score examples/writer/sharpening.md examples/llm-drafts/old-maps.md '<tmp>/writer.json' -o '<tmp>/mixed.json'
[exit 0]
--- stdout
STYLE COMPARISON   2 documents, 2 chunks, 1,257 words   vs writer.json (7 chunks)

By document   furthest from the reference first
  2 documents: 1 very different, 1 close
  document                      words  verdict             Delta  LLM-likeness
  examples/…rafts/old-maps.md     620  very different       4.65  like drafts
      most different in: Long words (7+ letters) ▲▲▲, List items ▲▲▲
  examples/…ter/sharpening.md     637  close                0.47  like reference

Across 2 documents: very different   Delta 2.56
  Lower is closer. The reference's own held-out writing at these lengths (620-637 words per chunk) scores 0.68 typically
  (a mean over 2 chunks, up to 0.87).
  Pooled over every document from here on; the table above has each one's own.

LLM-likeness: a few LLM traits   7.32
  The reference's own writing at these lengths (620-637 words per chunk) scores 0.20 typically (a mean over 2 chunks, up
  to 0.73); the LLM drafts 14.22.
  Strongest LLM signals: Long words (7+ letters) ▲▲▲, List items ▲▲▲, Bold phrases ▲▲▲

By area   Delta ÷ the top of the reference's usual range in each area; bars on a log scale
  close up to 1x, somewhat different to 1.5x, clearly different to 2x, very different above
  Formatting                  10.38x  ██████████████░░░░░░  very different
  Vocabulary                   5.17x  ██████████░░░░░░░░░░  very different
  Voice and stance             2.40x  ███████░░░░░░░░░░░░░  very different
  Punctuation                  1.39x  █████░░░░░░░░░░░░░░░  somewhat different
  Rhythm                       1.19x  ████░░░░░░░░░░░░░░░░  somewhat different
  Sentence shape               1.03x  ████░░░░░░░░░░░░░░░░  somewhat different
  Function words               0.95x  ████░░░░░░░░░░░░░░░░  close

Biggest differences   (each ▲ or ▼ is one standard deviation, up to 3)
                                                   this text   reference
  Long words (7+ letters)                                20%         11%   ▲▲▲
  List items                                         7.3 /1k     0.0 /1k   ▲▲▲  reference never varies; 1 of 2 chunks differ
  Bold phrases                                       7.3 /1k     0.0 /1k   ▲▲▲  reference never varies; 1 of 2 chunks differ
  Headings                                           7.2 /1k     1.6 /1k   ▲▲▲
  "these"                                            4.8 /1k     0.0 /1k   ▲▲▲  reference never varies; 1 of 2 chunks differ
  We / us / our                                      8.9 /1k     1.1 /1k   ▲▲▲
  Em dashes                                          3.2 /1k     0.0 /1k   ▲▲▲  reference never varies; 1 of 2 chunks differ
  Lexical diversity (MTLD)                               143        92.6   ▲▲▲

Pass --all to see every metric.

wrote <tmp>/mixed.json
