MCP Server implementations for managing context of students using a knowledge graph structure with entities, relations, and observations
Server has 6 tools with basic schemas and descriptions, but significant quality gaps. Tool names lack clear verb-noun structure (e.g., 'advanced_context' is vague). Descriptions are present but generic (10-50 chars for most tools). Parameter descriptions are minimal or absent. No output schemas documented. Error handling is not evident. The 'advanced_context' tool conflates 13 different operations into one tool, violating single-responsibility principle. Schemas are visible in code but incomplete, many parameters lack descriptions. No evidence of idempotency, pagination, or recovery guidance.
Advanced context operations including search, status tracking, and deadline management
Build and create new student context with entities and relations
Delete student context and related data
End the current session and save context
Load existing student context from memory
Start a new session for managing student context
advanced_context tool conflates 13 distinct operations (search_nodes, open_nodes, get_course_overview, get_upcoming_deadlines, etc.) into a single tool with an enum 'operation' parameter. This violates single-responsibility principle and forces LLMs to reason about which operation to select rather than calling a dedicated tool.
Tool descriptions are too short (10-50 chars) and lack context for LLM selection. E.g., 'Start a new session for managing student context' does not explain WHEN to call this vs load_context, what it returns, or what happens to existing sessions.
No output schemas documented. Tools return KnowledgeGraph objects but LLMs cannot see the structure of entities, relations, or observations fields. This forces LLMs to guess what data is available and how to extract it.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 42 | 2026-07-28+ | v2 |
Parameter descriptions are missing or trivial. E.g., 'observations' in build_context is described only as 'array of strings' with no guidance on what observations should contain or how they relate to entities.
No error handling guidance. Tools do not document what happens on invalid sessionId, missing entities, or constraint violations. LLMs have no recovery path if a call fails.