UnderTone
“When polite doesn’t mean optional.”
UnderTone separates how strongly a workplace request is phrased from the observable context that makes action expected.
Tuned for workplace English — Slack, Teams, and work email.
Example result — not computed on this page
Example result, not computed on this page — the real UnderTone output for the “Deadline + dependency” sample thread available on the Analyze page. Surface 3.1 out of 10, communicative force 6.8 out of 10, pragmatic gap +3.7, strongly under-phrased.
How it works
Surface
How strongly the request itself is phrased.
Signals: directness, modal wording, hedges, optionality formulas.
Communicative force
Observable surrounding language that makes action expected.
Signals: deadlines, consequences, dependencies, accountability, repetition.
Pragmatic gap
The difference between the two.
A large positive gap means the surrounding context carries more pressure than the request’s phrasing makes immediately visible.
Every score has a trace.
UnderTone highlights the precise spans of text that contribute to each score — nothing is scored without a visible reason.
Illustrative example only — not a live analysis. “Whenever you get a chance” is marked as surface evidence. “Legal is blocked on this” and “today” are marked as force evidence.
Local by design
UnderTone is an engineering instrument for workplace English, not an intent detector or a calibrated model of human perception.
Analysis runs entirely in your browser.
Message text is never sent to an UnderTone backend.
The diagnostic engine is deterministic and rule-based.