Loading page...Загружаем страницу...正在加载页面...
AtlasRepo

Designing a Website for Humans and AI

Description and use

A single website was built to serve executives, developers, content teams, and AI assistants using a component‑first CMS, Ghost‑to‑Contentful workflow, HubSpot forms, Markdown, OpenAPI, OAuth, and MCP.

Problem

CrewAI needed a single website that served executives, developers, content teams, and AI assistants without compromising usability for any group.

Outcome

The site now serves diverse human and AI consumers with a unified, structured content architecture, automated publishing, centralized forms, and multiple machine‑readable discovery surfaces.

Workflow

  • 1. Design component library in Figma
  • 2. Map components to Contentful content models
  • 3. Build GraphQL/TypeScript renderer
  • 4. Set up Ghost → Contentful automated transfer
  • 5. Centralize HubSpot forms with Zod validation
  • 6. Expose Markdown routes
  • 7. Publish OpenAPI, OAuth, MCP, and WebMCP endpoints
  • 8. Provide discovery surfaces (llms.txt, openapi.json, etc.)

Evidence

The project included 7 dedicated discovery surfaces for AI systems, a Contentful based content architecture, an automated publishing pipeline from Ghost to Contentful, centralized HubSpot form infrastructure, Markdown content delivery, OpenAPI documentation, OAuth discovery, and a Model Context Protocol (MCP) implementation.

Source

When an article is published in Ghost, an automated workflow converts it into a draft in Contentful.

Source

  • Record date: 2026-09-08T23:31:55.015Z