What is Brand Kit AI? Brand Kit AI is the set of AI powered features inside Brand Kit OS that populate, enrich, or generate brand data on your behalf.
It covers everything from extracting a full brand kit from a URL to enriching a target audience persona to drafting a blog post in your voice.
Every AI action in the app routes through the Lovable AI Gateway — a single provider agnostic boundary that lets us swap models per task without exposing provider keys to your browser.
The credit model 1 credit = 1 AI action unless the feature page says otherwise. Credits are deducted from the brand kit owner''s token pool. Members of a shared brand kit spend from the owner''s balance.
Failed runs (gateway errors, insufficient tokens, invalid input) do not consume credits. You can see your balance and per workflow history under Settings → Billing and, for admins, the AI Workflows dashboard at .
Where AI is used Every subsection covers one AI surface using the same template: What it does — the outcome you see Model — the exact model used today Prompt — a summary of the system prompt Fields written — the DB fields the workflow populates Cost — credits per run Where to trigger it — the UI entry point MCP equivalent — how to invoke the same capability over MCP, if available Failure modes — what to check when it doesn''t work Start with the pages linked in the Concepts sidebar: Extract from URL Target Audience enrichment Brand persona generation Competitor analysis Content & Generators Design principles 1.
Preview before write when possible. MCP exposes tools so agents propose changes before saving them. 2. Augment, don''t overwrite. Re running enrichment fills blanks; existing values win. 3. Owner pays.
Tokens come from the brand kit owner, regardless of who triggered the action. 4. Auditable. Every AI run writes to with model, tokens charged, duration, input, and output.