Introduction
Welcome to the Typograph Developer Documentation.
Typograph turns templates into press-ready PDFs and images: business cards, name badges, posters, invitations, direct mail. A designer builds the template and decides what may change in it. The people who fill it in personalise within those limits, whether they use a form on your site, your own system through the API, or an AI agent through the MCP server.
Who decides what can change
The template does. In the Editor, the designer gives each element a role: Fixed, Fill in the text, Place a picture, Fill in and place, Full edit, or a role of their own, with limits on how far an element may be moved or resized. The designer can also lock elements and choose which fields a form exposes.
Whichever route fills a template in, the print file comes from the same template and the same publisher.
The Viewer applies the template's roles and limits in the browser, and the Form Widget offers only the fields its form defines. Enforcing the same rules on the server for every API and MCP call is planned. Until then, a client with write access can change elements that a role protects, so give agents and integrations only the scopes they need.
What You'll Find Here
- Guides — step-by-step tutorials for integrating the browser scripts and the REST API
- Portal Documentation — end-user guides for the Typograph Portal
- OpenAPI Reference — interactive Swagger UI for the REST API
Platform Surface
Typograph exposes its capabilities through three complementary surfaces:
| Surface | Audience | Documented In |
|---|---|---|
| REST API | Any backend or client that can speak HTTP | Guides → API |
| Browser scripts | Web apps embedding the Editor, Viewer or Form Widget | Guides → Integration |
| MCP Server | AI agents (Claude, ChatGPT, Copilot, Cursor and others), acting for a signed-in user | AI Agents (MCP) |
The REST API and the MCP server both go through the same API Gateway, which handles OAuth 2.0 authentication, scope-based authorization, tiered rate limiting, and routing to the underlying services. The browser scripts run in your page and reach Typograph through the host providers you pass them. An agent connected through MCP works with the rights of the user who connected it.
Browser Scripts
The Editor and Viewer are delivered as CDN-hosted scripts — no npm package install is required.
| Script | Purpose | CDN URL |
|---|---|---|
| Editor | Full document editing UI | https://cdn.typograph.nl/editor/latest/editor.js |
| Viewer | Read-only template filling UI | https://cdn.typograph.nl/editor/latest/viewer.js |
| Form Widget | A form with a live preview, for end users | https://cdn.typograph.nl/form/latest/form.js |
The Editor and Viewer are constructed from the same bundle family and expose a window.typograph global, communicating with your page through DOM events and the host providers you pass at construction. See Editor Integration and Viewer Integration.
Quick Links
- Installation — create your first OAuth client and obtain credentials
- Quick Start — obtain a token and call the API
- Authentication — OAuth 2.0 flows (Client Credentials, Authorization Code + PKCE, Refresh)
- Core Concepts — the document model (templates, pages, elements, swatches)
- Support — contact, FAQ, status