Most tools follow a clear resource-oriented pattern (list_<entity>, get_<entity>), so list_students and get_student are easy to distinguish. However, a few overlapping concepts like coverage_tutors/coverage_summary/coverage_gaps and tutor_available_slots/list_availabilities could cause occasional misselection.

The dominant list_ and get_ prefix pattern is consistent and readable, but several tools deviate with noun-first or action-style names such as tutor_earnings, parent_balance, org_join_link, and coverage_summary. All names use snake_case, which keeps the overall set predictable despite the mixed conventions.

With 59 tools, this server far exceeds the reasonable scope for a focused MCP toolset and crosses the 50+ extreme threshold. Many tools are simple list/get pairs for niche resources like mileage, rate_cells, and blog_posts, making the set feel like an exhaustive read model rather than a curated tool surface.

The toolset is almost entirely read-only: there are no create, update, or delete operations for students, sessions, payments, invoices, or any other domain entity. This severely limits an agent's ability to perform basic tutoring-management workflows, leaving obvious lifecycle gaps.