How we gave our AI academy an MCP server agents can learn from
Описание и применение
Агенты Apex AI Academy получили доступ к образовательному контенту через ключ‑свободный JSON‑API и удалённый MCP‑сервер, позволяя автоматически получать уроки, курсы и новости, включая платный контент с персональными токенами.
Задача
Enabling AI agents to access and consume educational content from Apex AI Academy without a traditional web interface.
Результат
Agents can now programmatically access Apex AI Academy content, including premium lessons, via a stateless JSON‑RPC MCP server and a key‑free REST API, improving scalability and automation.
Рабочий процесс
- 1. Provide agents with three entry points: llms.txt guide, key‑free JSON API, and remote MCP server.
- 2. Agents call /api/public/v1/courses, /lessons/{slug}, /search, /news, /pricing via HTTP GET.
- 3. For agent‑specific interactions, agents use the MCP server at https://apexai.academy/mcp with JSON‑RPC over HTTP POST.
- 4. Agents invoke tools such as list_courses, get_lesson(view), search, get_pricing, and premium tools with bearer tokens.
- 5. Agents handle rate limits (60 req/min anon, 120 req/min key) and retry on 429s.
- 6. Agents can retrieve lesson views (agent, human, both) and send human‑friendly content to users.
- 7. The MCP server is registered in the official MCP Registry and discoverable via DNS TXT record.
- 8. Clients add the server to their mcpServers configuration to use it.
Доказательства
Three ways in for an agent: llms.txt, a key‑free JSON API, a remote MCP server at https://apexai.academy/mcp
Transport: Streamable HTTP, protocol 2025-06-18.
Tools. Public ones, no key: list_courses, list_lessons(course), list_deep_dives, get_lesson(slug, view), search(query), latest_news(limit 1–20), get_pricing, subscribe_link(plan?). With a Premium personal key: get_premium_lesson, get_progress, mark_lesson_complete.
We published to the official MCP Registry as academy.apexai/apex-ai-academy.
- Дата записи: 2026-10-17T03:56:02.655Z