AUDITED · FEB 4, 2026

Claude is a space to think

anthropic.com/news/claude-is-a-space-to-think
Measured in your browser by the same engine the MCP ships.
No server call. No tracking. Same code runs on any text.
D score
0.82
elevated
narrative
0.45
engagement
0.09
mimicry
0.65
01 — measurement
tokens1,162
sentences57
vocabulary entropy9.53 bits
lexical diversity0.424
hedge density0.09%
meta-commentary density0.17%
avg sentence length20.4 tokens
stopword ratio0.441
bigram repetition0.215
contains citationsno
assertion-bearing sentences27 of 57
flagged (claim, no cite)27
02 — reading A high D driven by the mimicry valve, not the narrative valve. The post makes 27 assertion-bearing statements across 57 sentences and contains zero citations, URLs, or "according to" references. That's the pattern the engine is responding to — not that the piece is performed, but that every claim is delivered as self-evident. It's a policy declaration, and policy declarations from AI labs are exactly the content type that benefits from citation discipline. The engine is not saying the piece is wrong; it's saying the piece is unsupported by design.