No user-submitted related servers found.
Scored across 32 tools
The toolset spans distinct stages, but several read/query tool families overlap significantly (read_document vs read_element vs read_evidence vs read_element_evidence; read_practice_feedback vs read_practice_history). Their descriptions are detailed and help, yet the sheer number of superficially similar 'read_*' and 'list_*' tools creates real misselection risk.
Almost all tools use snake_case with a verb_noun pattern (import_document, read_study_record, update_review_plan), which is predictable. Minor deviations like workspace_status (noun phrase) and compound names such as resume_job_observation keep it from perfect consistency.
32 tools far exceed the typical well-scoped range of 3–15 and land in the 'too many' bucket per the rubric. Although each tool maps to a real sub-operation in a complex study-errorbook workflow, the volume imposes substantial cognitive load and coordination burden.
The surface covers import, OCR parsing, evidence extraction, study-record CRUD, practice recording/history, review planning, and export—a broad lifecycle. Only minor gaps remain, such as explicit parse-job cancellation or bulk study-record management, which agents can likely work around.