Skip to content

Developers

CourseAdvisor for developers and AI agents

You can read every page as markdown, ask our MCP server about products and articles, and send a demo enquiry for a user. Nothing here needs an API key.

Pages as markdown

Send Accept: text/markdown to any page URL and you get markdown back with Content-Type: text/markdown. Browsers still get HTML. You can also fetch /<path>/index.md directly.

curl -s https://courseadvisor.ai/pricing -H "Accept: text/markdown"

MCP server

Endpoint: https://courseadvisor.ai/mcp. Transport: Streamable HTTP, stateless, JSON responses. No auth.

  • list_products: every product with a summary and page URL.
  • get_product: how one product works, who it fits, its pricing models and FAQs.
  • list_articles: published blog articles, newest first.
  • get_page: any page on this site as markdown.
  • request_demo: sends a demo enquiry to our team, with the same checks as the contact form. Use it only when the user asks.
curl -s https://courseadvisor.ai/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"list_products","arguments":{}}}'

Public API

POST /api/contact takes JSON or form data with name, email, organisation and message, plus optional role and leadVolume. It allows 5 submissions per IP every 10 minutes. GET /api/health reports status. The full schema is in openapi.json.

curl -s https://courseadvisor.ai/api/contact \
  -H "Content-Type: application/json" \
  -d '{"name":"Alex Chen","email":"alex@provider.edu.au","organisation":"Example College","message":"We get 3,000 enquiries a month and call most of them back the next day."}'

In the browser (WebMCP)

Pages register WebMCP tools when the browser supports navigator.modelContext: list_products, open_page and request_demo. The contact form also carries a declarative toolname attribute.

Discovery files

Questions? Email hello@courseadvisor.ai.