Brand Atlas

Connect

Every AI gets the same context.

Authorize once via MCP, the REST API, or the Gateway. ChatGPT, Claude, Cursor and your stack all read the same Brand, Products and Claims — not a pasted PDF and not a different prompt in every tool.

  • ClaudeClaude
  • ChatGPTChatGPT
  • CursorCursor
  • Claude CodeClaude Code
  • VS CodeVS Code
  • CodexCodex
  • LovableLovable
  • MCP
  • REST API
  • Gateway

How it works

One Atlas. Every interface.

  1. 01

    Build the Atlas

    Brand, Products and Claims become the source. Connect has nothing to send until they exist.

  2. 02

    Authorize once

    MCP, the REST API, or the Gateway. One connection — scoped to a brand or the workspace.

  3. 03

    Same packet

    Every request pulls Brand, the real SKU, and the approved claims. Not a different brief in each tool.

  4. 04

    Ask from anywhere

    ChatGPT, Claude, Cursor, Claude Code, VS Code, Codex, Lovable, your own app, or traffic through the Gateway. Same answer.

Why it matters

Same context. Every AI.

01

Same context everywhere

The Brand Book, the catalog and the policy do not fork when someone opens a new chat.

Where it shows up

Same jobs. Every interface.

Not a shortlist of tools. Ask these in Claude, ChatGPT, Cursor, VS Code, Lovable — over MCP, the REST API, or the Gateway.

01 · Policy

Ask what you can say

“What can I say about Focus in Austria?” returns the approved line — not a stronger one the model invented.

Why it exists

Not a prompt you paste into every tool.

Tribal knowledge and exported PDFs drift the moment someone opens a new chat. One packet does not.

Without Connect

A different brief in every interface.

  • Someone pastes last quarter’s PDF.
  • Claude knows a claim Cursor never saw.
  • The SKU in ChatGPT is a lookalike.

With Connect

One context packet. Every AI.

  • Brand, Products and Claims stay the source.
  • Authorize once — MCP, API or Gateway.
  • The same SKU and policy, whatever the interface.

Build your Atlas

Connect has nothing to send until the Atlas exists.

Brand, Products and Claims are the packet. Compile them first.

Brand

A brand book AI can actually use.

Identity, colour, type, voice and knowledge — compiled once, then kept editable.

Learn more

Products

The SKU AI can actually use.

The right SKU, packshot, on-pack spelling and variant — not a lookalike from a prompt.

Learn more

Claims

Your policy, enforced.

Approved claims, evidence and rules travel with every relevant request.

Learn more

Then put it to work

Create. Check.

Create when you want stills from the catalog. Check before anything ships. Both read the same source Connect just distributed.

Create

Create without losing control.

Turn one verified product into channel-ready creative — without losing the brand, product or approved claims.

Learn more

Check

Check before you ship.

Brand Match on Create. Claims Check via gateway, API and MCP. Not a universal output firewall.

Learn more

Give every AI the same context.

Questions

Does it work with the AI we already use?

Yes. ChatGPT, Claude, Cursor and your other tools can use the same Brand, Products and Claims. Authorize once via MCP, or point an existing OpenAI, Anthropic or Gemini client at the Gateway.

How does Connect work?

Authorize once via MCP, the REST API, or the Gateway. ChatGPT, Claude, Cursor, VS Code, Lovable and your own apps — including OpenAI, Anthropic or Gemini clients you already call — receive the same Brand, Products and Claims.

Do we have to generate in Brand Atlas?

No. Create is there when you want on-brand stills from the catalog. If you keep writing and designing in ChatGPT or Cursor, Connect and Check still apply.

Does it watch every AI the team uses?

No. It is not a live monitor of every model or chat. Check runs when you create in Brand Atlas, or when a request goes through the gateway, API or MCP.

Can we use it with multiple brands?

Yes. Each workspace can hold multiple brands. MCP connections can be scoped to a brand or the organization.