Brand kit, voice and rules

ReBrandOS is the source of truth your AI client reads before it writes. Set your brand up once, and every conversation starts from the same voice, colours and rules.

Where to set it up

Open Brand setup for the brand, then Identity:

• Brief & strategy: voice traits, what to avoid sounding like, typography, whether content may name clients and companies, and the clients to keep private.
• Look & feel: primary, secondary and accent colours, and the style category.

Your AI client can also read and update the brand through the connector, including adding a new brand kit version. Each version keeps its colours, typography, style, voice and template settings, so you can see what changed.

Voice and words to avoid

A brand’s voice can include traits, do’s and don’ts, words the brand never uses, call-to-action lines, signature lines and an emoji policy. Ask your AI to add words to avoid, for example:

Add “guaranteed” and “best ever” to Northwind Coffee’s words to avoid.

When you approve posts, ReBrandOS keeps your five most recent approved texts as voice samples, and your AI reads them with the rest of the brand guidance.

How the rules are checked

• Protected client names block a draft. While a brand does not allow naming clients, a draft or slide that names one of its private clients is refused, with every match listed and generic wording suggested.
• Other rules are warnings. Words to avoid, emoji when the brand’s policy is none, and more than five hashtags are flagged so you or your AI can fix them before you approve.

Checks run when a draft is previewed, saved or updated, and on carousel slide text.