The search takes numeric subject, grade and feature ids rather than names, which would normally mean digging through a site to find them; the three master tools are what make it usable from a conversation. Sorting by rating or lesson count is where the real filtering happens once a search returns more tutors than anyone wants to read.
A Node MCP server for Manalink, an online tutoring platform. It exposes the master lists you need to build a search, plus the search itself.
- `get_subject_master` — subject names paired with the ids the search expects
- `get_grade_master` — grade names and ids
- `get_course_features` — the course feature master, fetched live from the API
- `search_teachers_advanced` — search by `subject_ids`, `grade_ids` and `course_feature_id`, with `sort` over pr, certification, rating, lesson count or latest, and a teaching-period filter
- Results come back as HTML and as Markdown, so the assistant can read the clean version
Node.js 18+ and npm or yarn. Clone, `npm install`, `npm run build`, then point Claude for Desktop at the built dist/index.js with an absolute path and restart it.
Build from source — clone the repository and build it, then point your client at the binary
