middleschool-tutor-gql

Exposes middle school curriculum content (math, science, language arts, social studies) through a GraphQL API schema, allowing clients to query structured educational materials with field-level granularity. Implements resolver functions that fetch or generate tutoring content based on query parameters like subject, grade level, and topic, enabling dynamic content retrieval without fixed REST endpoints.

Implements the Model Context Protocol (MCP) server specification, exposing educational content tools as MCP resources and tools that Claude or other MCP-compatible LLMs can discover and invoke. Handles MCP protocol handshake, resource listing, tool schema advertisement, and request/response serialization, allowing AI agents to treat curriculum queries as native capabilities.

Resolves educational content queries by mapping subject names (math, science, language arts, social studies) and topic hierarchies (e.g., algebra > linear equations > solving for x) to structured curriculum data. Uses resolver functions to fetch or generate explanations, examples, and practice problems based on grade level and difficulty parameters, supporting multi-level topic nesting.

Maintains conversation state across multiple tutoring interactions by leveraging MCP's context protocol, allowing the server to track student progress, previous questions, and learning history within a single tutoring session. Resolvers can access prior query context to provide personalized follow-up content and avoid repeating explanations.

Generates detailed worked examples for math and science problems by breaking solutions into discrete steps with explanations at each stage. Implements a resolver that structures problem-solving workflows (e.g., 'identify given', 'set up equation', 'solve', 'verify') and provides reasoning for each step, enabling students to learn problem-solving methodology alongside content.

Generates practice problems for middle school subjects with corresponding answer keys and difficulty levels calibrated to grade and topic. Implements resolvers that create problem variants (e.g., different numbers, contexts) from templates and assign difficulty scores based on cognitive complexity, enabling adaptive problem sequencing.

Maps curriculum content to grade levels (6-8) and learning standards (e.g., Common Core, state standards) through metadata resolvers that tag topics with standard codes and grade appropriateness. Enables queries filtered by grade level or standard, allowing educators to ensure content aligns with curriculum requirements.