A stateless MCP server for Google Classroom — read courses, grade submissions, and auto-submit assignments.
Server has 11 tools with basic schemas and descriptions, but significant gaps limit production readiness. All tools have descriptions (10 - 200 chars, mostly adequate), and input schemas are present with type definitions. However, parameter descriptions are minimal or missing context about constraints, dependencies, and error recovery. Output schemas are undocumented, LLMs cannot predict response structure. No tool annotations (readOnlyHint/destructiveHint) despite clear risk levels. Error handling is absent; tools fail silently or return raw API errors. Naming is verb-forward (list_, grade_, submit_) but lacks clarity on mutual exclusivity and chaining requirements. No pagination guidance for list tools returning potentially large datasets.
Grade a student's submission for a specific coursework assignment using an LLM. Args: course_id: The unique Google Classroom Course ID. coursework_id: The unique Google Classroom Coursework ID. student_email: The student's email address.
List all announcements in a Google Classroom course. Args: course_id: The unique Google Classroom Course ID.
List all active Google Classroom courses where you are a teacher or student. Returns a JSON array of courses with their IDs, names, sections, and links.
List all coursework (assignments) registered in a Google Classroom course. Args: course_id: The unique Google Classroom Course ID.
List all students enrolled in a specific Google Classroom course. Args: course_id: The unique Google Classroom Course ID.
Output schemas completely undocumented. LLMs cannot predict response structure (fields, types, pagination). Forces agents to guess or fail on parsing.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Tools marked READ_ONLY or WRITE in metadata but not exposed to LLM via MCP annotations. Agents cannot distinguish safe from destructive calls.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
List all student submissions for a specific coursework assignment. Args: course_id: The unique Google Classroom Course ID. coursework_id: The unique Google Classroom Coursework ID.
List all teachers in a specific Google Classroom course. Args: course_id: The unique Google Classroom Course ID.
Update the grade for a student submission. Args: course_id: The unique Google Classroom Course ID. coursework_id: The unique Google Classroom Coursework ID. student_email: The student's email address. grade: The numeric grade to assign.
Return a graded submission to a student (mark as returned in Google Classroom). Args: course_id: The unique Google Classroom Course ID. coursework_id: The unique Google Classroom Coursework ID. student_email: The student's email address.
Submit a solution to a coursework assignment on behalf of the authenticated student. Args: course_id: The unique Google Classroom Course ID. coursework_id: The unique Google Classroom Coursework ID. submission_text: The text content of the submission.
Turn in (finalize) a student submission for a coursework assignment. Args: course_id: The unique Google Classroom Course ID. coursework_id: The unique Google Classroom Coursework ID.
Parameter descriptions lack constraint details. 'course_id' and 'coursework_id' described only as 'The unique Google Classroom Course ID', no format, length, or validation rules. LLMs cannot self-correct invalid input.
No error handling or recovery guidance. Tools fail silently or return raw Google API errors. No actionable error messages to guide LLM retry or fallback logic.
List tools lack pagination guidance. No limit, offset, or cursor parameters visible. Large result sets (e.g., 500 students) will blow context window and degrade LLM reasoning.