Voice and tone - Invoca

See Voice and tone for the decided part of this page.

Gaps in the current rules

Gap	Why it matters	Status
No mapping from brand personality to interface copy. The three traits describe the brand; nothing in the source says how “Eye-opening” or “Bold” should show up in a button label, an error message, or an empty state.	Without it, two writers can both claim to be “on brand” while producing very different copy for the same screen.	Open
No worked examples. The source states the traits but does not show a “sounds like us” / “doesn’t sound like us” comparison for any real screen.	Traits without examples are hard to apply consistently across a team.	Open
Possible tension with caution-first patterns. “Bold” — assertive, willing to take risks — is a reasonable fit for marketing headlines and a much less obvious fit for, say, a destructive-confirmation dialog, where the system’s own decision is to be cautious rather than assertive (see Destructive confirmation). Not reconciled against places the system deliberately asks for restraint.	An untested tension is one a writer discovers by shipping the wrong tone, not by reading the docs.	Open

Constraints Titan should enforce

These are real, currently-true rules. Nothing in code stops a builder from violating them today — each is a placeholder for a guardrail Titan hasn’t built, not a decision still being made.

ID	Constraint	Would be enforced by
TITAN-VOICE-01	Product copy uses plain language over technical or internal wording.	A content linter (e.g. a Vale-style rule) checking copy against a maintained list of internal or technical terms.
TITAN-VOICE-02	Avoid acronyms and phrases not commonly heard outside the office.	The same content linter, checking copy against a maintained acronym/jargon list.

TITAN-VOICE-03 (“concise and clear”) is excluded above — conciseness isn’t mechanically checkable, so the only guardrail available today is editorial review, not a lint rule.

Last modified on September 2, 2026