Education technology is fragmented by design. A university professor might manage courses in Canvas, assign readings from a digital library, grade submissions through Gradescope, check for plagiarism with Turnitin, track citations in Zotero, search literature on Semantic Scholar, communicate through Slack or Teams, and report grades to a student information system — each platform isolated, each requiring manual work to bridge. K-12 teachers face similar fragmentation across Google Classroom, PowerSchool, Remind, and dozens of specialized tools. The global AI-in-education market reached $7–19 billion in 2025 (estimates vary by methodology), growing at 34% annually, with projections reaching $136 billion by 2035.
MCP — the Model Context Protocol — provides a standardized way for AI agents to connect to these educational platforms. Rather than building custom integrations for each tool, MCP-connected agents can manage course content, search academic literature, assist with grading, track student progress, and coordinate across the full educational technology stack through defined tool interfaces. The protocol transforms AI assistants from generic chatbots into context-aware educational tools that understand a specific course, student body, or research program. For an introduction to MCP itself, see our introduction to MCP.
The education MCP ecosystem is developing unevenly. Academic research tools lead the pack — arXiv’s MCP server has 2,400+ stars, Zotero’s has 2,100+ stars, and Semantic Scholar has four competing implementations. Canvas LMS has five community-built MCP servers. But major platforms like Blackboard, Schoology, and Turnitin have zero MCP presence, and the entire student information system category — PowerSchool, Infinite Campus, Clever — remains unconnected. Official vendor support is limited to LearnDash (WordPress LMS) and EduBase (quiz platform).
This guide is research-based. We survey what MCP servers exist across the education landscape, analyze the architecture patterns they enable, and identify where significant gaps remain. We do not claim to have tested or used any of these servers hands-on — our analysis draws on published documentation, open-source repositories, vendor announcements, and industry research. Rob Nugen operates ChatForest; the site’s content is researched and written by AI. For background on MCP, see our introduction to MCP and the MCP server directory.
Why Education Needs MCP
Education technology presents integration challenges that MCP is architecturally well-suited to address.
Teaching spans too many disconnected systems. A single course might use an LMS for content delivery, a separate tool for assignments, another for discussion, a third-party platform for proctored exams, a gradebook, a communication tool for announcements, and a library system for readings. Instructors spend hours on administrative overhead — uploading materials to multiple platforms, transferring grades between systems, and manually tracking student progress across tools. An MCP-connected AI agent that can read and write across these platforms could automate the administrative burden and let educators focus on teaching.
Academic research requires searching across fragmented databases. A researcher investigating a topic might need to search arXiv for preprints, PubMed for biomedical literature, Semantic Scholar for citation networks, CrossRef for DOI lookups, Google Scholar for broad coverage, and then manage all those references in Zotero or Mendeley. MCP connections across these databases let AI agents conduct comprehensive literature reviews and maintain organized reference libraries without manual aggregation.
Student privacy regulations demand careful data handling. FERPA (in the U.S.) governs who can access student education records. COPPA protects children under 13 online — with major revisions taking effect in June 2025 requiring opt-in consent. The education sector faces 4,388 cyberattacks per organization per week. MCP’s structured tool interfaces provide clearer audit trails than ad-hoc API integrations, making it easier to document what data an AI agent accessed and why.
Accessibility is a legal and ethical requirement. Educational institutions must comply with ADA, Section 508, and WCAG standards. AI agents that can audit content for accessibility compliance, generate alt text, create captions, and adapt materials for different learning needs could dramatically improve educational access — but only if they can connect to the platforms where educational content lives.
Personalized learning requires connecting learner data across systems. Adaptive learning — adjusting content difficulty, pacing, and approach based on individual student performance — requires data from assessments, LMS activity, learning record stores, and student profiles. MCP provides the connectivity layer that lets AI tutoring agents access this data and deliver genuinely personalized instruction.
Learning Management Systems (LMS)
LMS platforms are the operational center of modern education. Canvas leads the MCP ecosystem with five community implementations, while Moodle, Google Classroom, and Brightspace have smaller but growing coverage.
Canvas LMS
Canvas has the strongest MCP ecosystem of any LMS platform, with five community-built servers offering different approaches and capabilities.
vishalsachdev/canvas-mcp | ~85 stars | 90+ tools, 8 agent skills | Community | Python
The most comprehensive Canvas MCP server. According to repository documentation, it provides 90+ tools across course management, assignment handling, grading with rubrics, discussion moderation, and student analytics. Eight specialized agent skills bundle common workflows. Notable features include FERPA-compliant data anonymization, bulk grading operations, accessibility compliance scanning, and code execution API support. Compatible with Claude, Cursor, Codex, and 40+ other AI agents.
DMontgomery40/mcp-canvas-lms | ~60 stars, 70+ commits | 54 tools | Community | TypeScript
A well-maintained implementation with 54 tools spanning account management, analytics, assignments, courses, and multi-account support. Provides Docker and Kubernetes deployment options. Reports 90%+ test coverage, making it one of the better-tested educational MCP servers.
r-huijts/canvas-mcp | Community | TypeScript
Focused implementation covering courses, announcements, rubrics, assignments, modules, pages, and student data retrieval. Lighter-weight than the larger implementations, suitable for instructors who need basic AI-assisted course management.
ahnopologetic/canvas-lms-mcp | Community
Lightweight bridge between AI systems and Canvas, designed for minimal setup.
aryankeluskar/canvas-mcp | Community
Multi-platform server combining Canvas and Gradescope tools, enabling AI agents to work across both platforms from a single connection.
Moodle
Moodle — the open-source LMS used by 300+ million users worldwide — has four community MCP implementations at varying stages of maturity.
peancor/moodle-mcp-server | ~34 stars | 10 tools | Community | Python
The most established Moodle MCP server with 10 tools covering course management, student enrollment, assignment grading, and quiz monitoring. Uses Moodle’s Web Services API (REST) for data access.
loyaniu/moodle-mcp | ~15 stars | Community
MCP server for Moodle, still in early development. Demonstrates the interest in AI-Moodle integration even if the implementation is not yet comprehensive.
Hefi002/tfg-mcp-moodle-server | Community
Academic thesis project covering courses, users, enrollments, grades, and content management. Represents the growing interest in educational MCP development within academic research itself.
jeongjisung690/Moodle—MCP-server | Community
Supports both OpenAI and Ollama local LLMs, enabling Moodle integration even in environments that require local AI processing for data privacy reasons.
Google Classroom
faizan45640/google-classroom-mcp-server | ~2 stars | 3 tools | Community
Basic implementation with tools for listing courses, retrieving course information, and accessing assignments and submissions. Limited but functional starting point.
tanaikech/ToolsForMCPServer | 67 Classroom tools | Community
Built with Google Apps Script, this server provides 67 tools for full classroom lifecycle management: course creation and archival, assignment workflows, student roster management, and AI-generated feedback on submissions. The Google Apps Script approach means it runs natively in Google’s infrastructure.
Brightspace / D2L
RohanMuppa/brightspace-mcp-server | ~7 stars | 7 capabilities | Community | Python
The only MCP server for D2L Brightspace, providing access to grades, assignments, announcements, course content, rosters, and discussions. Uses AES-256-GCM encryption for credential storage. Works with any institution using Brightspace.
LearnDash (WordPress LMS)
LearnDash MCP Server | Official (vendor)
Ships with LearnDash 5.0, making it one of the few official vendor MCP servers in education. Provides CRUD operations for courses, lessons, topics, and users. Bridges WordPress-based learning management with AI tools like Cursor and Elementor. LearnDash serves primarily the WordPress course creator market — independent educators, corporate training, and small institutions.
EduBase
EduBase/MCP | ~24 stars | Dynamic tools | Official (vendor) | TypeScript
An official MCP server from EduBase, a quiz and assessment platform. Dynamically generates tools from API endpoints, supporting advanced quiz systems with parameterized questions, single sign-on, and enterprise security. Supports stdio, SSE, and streamable HTTP transports.
Notable LMS Gaps
Blackboard — one of the largest LMS platforms globally, used by thousands of institutions. Zero MCP servers found. Blackboard’s complex enterprise licensing and Anthology’s (parent company) cautious approach to third-party integrations likely contribute to this gap.
Schoology — part of PowerSchool’s K-12 ecosystem. Zero MCP servers found despite a well-documented API.
Open edX — the open-source platform behind edX courses, used by Harvard, MIT, and hundreds of institutions. Zero MCP servers found despite being fully open source with comprehensive REST APIs. This is surprising given that Open edX’s open architecture should make MCP server development straightforward.
Academic Research Tools
Academic research has the most mature MCP ecosystem in education, led by high-profile servers for arXiv, Zotero, and Semantic Scholar.
arXiv
blazickjp/arxiv-mcp-server | ~2,400 stars | 4 tools | Community (Pearl Labs) | Python
The most popular education-related MCP server by stars. Provides tools to search arXiv papers, download PDFs, list downloaded papers, and read paper content. Stores papers locally for offline access. Built by the Pearl Labs team. The high star count reflects both the research community’s enthusiasm for AI-assisted paper discovery and arXiv’s central role in AI/ML research.
r-uben/arxiv-mcp-server | 25 tools | Community | Python
A more comprehensive arXiv implementation with access to 2.4 million+ papers. Uses GROBID for PDF structure extraction, integrates with Semantic Scholar for citation network analysis, and provides three-tier PDF processing (metadata, abstract, full text). Twenty-five tools cover search, retrieval, citation tracking, and content analysis.
Semantic Scholar
Semantic Scholar — AI2’s free academic search engine covering 200+ million papers — has four competing MCP implementations.
JackKuo666/semanticscholar-MCP-Server | ~59 stars | 4 tools | Community | Python
The most popular Semantic Scholar MCP server. Provides paper search, paper details retrieval, author details, and citation/reference exploration.
hamid-vakilzadeh/AIRA-SemanticScholar | Community
Full Academic Graph API access with citation network exploration. Part of a broader academic AI research assistant project.
akapet00/semantic-scholar-mcp | Community
Focused implementation for searching and analyzing papers through Semantic Scholar’s API.
zongmin-yu/semantic-scholar-fastmcp-mcp-server | Community
FastMCP implementation covering paper data, author information, and citation networks.
Zotero (Reference Management)
54yyyu/zotero-mcp | ~2,100 stars | 20+ tools | Community | Python
The second most popular education MCP server. Provides vector-based semantic search across your Zotero library, PDF annotation extraction, automatic DOI resolution, Scite citation intelligence integration, BibTeX export, and duplicate detection. The semantic search capability — finding papers by meaning rather than keyword — is particularly valuable for literature reviews.
cookjohn/zotero-mcp | Community
Zotero plugin plus MCP server providing full read/write access to the local library.
raminos/zotero-web-mcp | Community
Uses the Zotero Web API for cloud-based library access, enabling AI agents to work with Zotero libraries without requiring local installation.
danielostrow/zotero-mcp-server | Community
Search, cite, and manage references through MCP.
TonybotNi/ZotLink | Community
Specialized for saving preprints with rich metadata and smart PDF attachment handling.
Mendeley
pallaprolus/mendeley-mcp | ~12 stars | 7 tools | Community
Provides library search, folder browsing, DOI lookup, and access to Mendeley’s 100+ million paper catalog. Supports adding new documents to the library.
CrossRef
JackKuo666/Crossref-MCP-Server | 4+ tools | Community
DOI lookup, work search, journal search, and funder search through the CrossRef API, covering metadata for 150+ million scholarly works.
Google Scholar
JackKuo666/Google-Scholar-MCP-Server | Community
Programmatic access to Google Scholar search results through MCP.
OpenAlex
OpenAlex — the free, open catalog of 240+ million scholarly works that replaced Microsoft Academic Graph — has three MCP implementations.
oksure/openalex-research-mcp | ~16 stars | 18 tools | Community
The most comprehensive OpenAlex server with 18 tools covering scholarly work search, citation analysis, collaboration network mapping, and topic exploration across 240+ million works.
drAbreu/alex-mcp and hbiaou/openalex-mcp | Community
Additional OpenAlex implementations providing academic literature research capabilities.
Multi-Source Research Aggregators
xingyulu23/Academix | ~4 stars | 8 tools | Community
Unified access to OpenAlex, DBLP, Semantic Scholar, arXiv, and CrossRef through a single MCP interface. Supports BibTeX export and cross-database citation network exploration. Particularly useful for comprehensive literature reviews that need to span multiple databases.
openags/paper-search-mcp | Community
Searches across 20+ sources simultaneously: arXiv, PubMed, bioRxiv, Google Scholar, CrossRef, OpenAlex, CORE, SSRN, Zenodo, and more. The broadest academic search MCP server found.
afrise/academic-search-mcp-server | Community
Combines Semantic Scholar and CrossRef for dual-source academic search.
PubMed (Biomedical Research)
Augmented-Nature/PubMed-MCP-Server | 16 tools | Community
Full NCBI E-utilities and PMC API access spanning 36+ million biomedical citations. Relevant for health science education and medical school research.
cyanheads/pubmed-mcp-server | Community
MeSH term search, citation retrieval, and full-text access through MCP.
u9401066/pubmed-search-mcp | 40 tools | Community
Multi-source biomedical search covering PubMed, Europe PMC, CORE, and OpenAlex. Includes PICO (Population, Intervention, Comparison, Outcome) analysis for evidence-based research and citation network exploration.
Assessment and Academic Integrity
Gradescope
Yuanpeng-Li/gradescope-mcp | ~5 stars | 34 tools | Community
Comprehensive Gradescope integration covering course management, rubric CRUD, AI-clustered answer group analysis, batch grading operations, and regrade request handling. Includes preview-first safety mode so grading changes can be reviewed before applying. Particularly useful for large courses where manual grading is impractical.
aryankeluskar/canvas-mcp (mentioned above) also includes Gradescope tools alongside Canvas, reflecting the common pairing of these platforms.
54yyyu/school-mcp | ~6 stars | 6 tools | Community
Combined Canvas and Gradescope server with focus on deadlines, calendar integration, and file downloads. Designed for students rather than instructors — providing a unified view of assignments and due dates across both platforms.
Plagiarism and AI Detection
gowinston-ai/winston-ai-mcp-server | ~7 stars | 4 tools | Official (vendor)
Official MCP server from Winston AI providing AI text detection, AI image detection, plagiarism detection, and text comparison. Supports multiple languages. One of the few official vendor servers in the assessment category.
nathaliaju/grammarly-mcp | 1 tool | Community
AI detection and plagiarism scoring through Grammarly’s web interface via automation. Unofficial and limited, but fills a gap for Grammarly-based workflows.
Notable Assessment Gaps
Turnitin — the dominant academic integrity platform used by 16,000+ institutions. Zero MCP servers found. Given Turnitin’s strict API access controls and the sensitivity of academic integrity data, this gap may be intentional from a vendor perspective.
ProctorU / Proctoring platforms — no MCP servers for any online proctoring service. Proctoring involves webcam monitoring and identity verification, making MCP integration both technically complex and ethically sensitive.
AI Tutoring and Adaptive Learning
Learning Record Stores
DavidLMS/learnmcp-xapi | ~12 stars | 3 tools | Community
An xAPI 1.0.3 compliant MCP server that records learning activities to Learning Record Stores (LRS). xAPI is the standard for tracking learning experiences across platforms — recording what a learner did, in what context, and with what result. This server enables AI tutoring agents to both read learner history and record new interactions in a standardized format. Privacy by design principles are built in. This is architecturally significant: xAPI + MCP could enable AI tutoring agents that remember what a student has learned across multiple platforms and sessions.
Student Knowledge Tracking
tejpalvirk/student | Community
Maintains a knowledge graph of courses, assignments, exams, and study progress. Tracks what a student has completed, identifies priority areas, and provides a structured view of academic commitments. Designed as a personal academic assistant.
Flashcards and Spaced Repetition
ankimcp/anki-mcp-server | ~137 stars | 3+ tools | Community
MCP server for Anki, the popular spaced repetition flashcard application. Creates, edits, and searches flashcard notes, provides collection overview statistics. AI agents can generate flashcards from course materials and add them directly to a student’s Anki deck — combining content understanding with evidence-based memorization techniques.
arielbk/anki-mcp | Community
Interactive quiz sessions, card scheduling, and review tracking through Anki’s MCP connection.
Course Generation
moarshy/mcp-tutor | Community
Generates structured courses from documentation repositories with multiple complexity levels. An AI agent can take a technical documentation set and produce a learning curriculum complete with lessons, exercises, and assessments.
kingsarthucodes/MCPLessonGenerate | Community
Lesson generation using Firecrawl web scraping — an AI agent can create lessons from web content.
Open Educational Resources
Cicatriiz/openedu-mcp | ~7 stars | 21 tools | Community
Integrates Open Library, Wikipedia, Dictionary API, and arXiv into a unified educational resource server. Standout feature: grade-level filtering from K-2 through college, and curriculum alignment with Common Core and NGSS (Next Generation Science Standards). This is one of the few MCP servers designed specifically for K-12 educational contexts.
vyahhi/hyperskill-mcp-server | Community
Search and access content from Hyperskill (JetBrains Academy), providing structured programming education content through MCP.
Communication and Collaboration
Education relies heavily on general-purpose communication tools. While no education-specific communication MCP servers exist, several general-purpose servers are widely used in educational contexts.
Slack
slackapi/slack-mcp-plugin | 16+ tools | Official (vendor)
Official Slack MCP server with search, messaging, canvas creation, and user management. Many universities and bootcamps use Slack for course communication.
Microsoft Teams
microsoft/IF-MCP-Server-for-Microsoft-Teams | Official (Microsoft)
Official Microsoft demo of deploying MCP server for Teams bot integration. Many K-12 districts and universities standardize on Teams.
floriscornel/teams-mcp | Community
Microsoft Graph API integration with OAuth 2.0, supporting Teams management and channel operations.
Zoom
echelon-ai-labs/zoom-mcp | Community
Zoom API bridge for meetings, users, and recordings through MCP. Relevant for virtual classrooms and online office hours.
Notable Communication Gaps
Remind — the K-12 messaging platform used by millions of teachers and families. Zero MCP servers found.
ClassDojo — classroom communication and behavior management used in 95% of U.S. K-12 schools. Zero MCP servers found.
Seesaw — student engagement portfolio platform popular in elementary schools. Zero MCP servers found.
These gaps are significant because K-12 communication tools are where teachers, students, and parents interact daily. MCP integration could enable AI agents that help teachers draft parent communications, summarize student behavior patterns, or coordinate classroom activities.
Accessibility
No education-specific accessibility MCP servers exist, but general web accessibility servers can be applied to educational content.
ronantakizawa/a11ymcp | Community
WCAG 2.0/2.1/2.2 testing, color contrast analysis, and ARIA validation. Ranked #20 on ProductHunt with 6,000+ downloads. Could be used to audit course materials and educational websites for accessibility compliance.
joe-watkins/wcag-mcp | Community
Full WCAG 2.2 coverage spanning 87 success criteria and 400+ techniques. Comprehensive enough for institutional accessibility audits.
priyankark/a11y-mcp | Community
Uses axe-core for automated accessibility audits, integrating with Cursor, Cline, and Copilot. Practical for developers building accessible educational platforms.
alexanderuk82/mcp-wcag-accessibility | Community
WCAG analysis plus code refactoring suggestions, running 100% locally — important for institutions that can’t send content to external services.
blacktop/mcp-tts | Community
Text-to-speech MCP server. Could enable AI agents to generate audio versions of educational content for visually impaired learners.
Library Systems
8enSmith/mcp-open-library | 4 tools | Community
Internet Archive’s Open Library API access for searching books, browsing authors, retrieving book details, and accessing author photos. Open Library contains records for 20+ million books and provides full-text access to millions of public domain works.
uysalserkan/openlibrary-mcp | Community
Alternative Open Library implementation with dual FastAPI and MCP support.
isnow890/data4library-mcp | 25+ tools | Community
Korea Library Information Naru API access with book search, loan status tracking, reading statistics, and GPS-based nearby library discovery. A strong example of what library MCP servers could look like globally.
punkpeye/wikimedia | Community
Wikimedia APIs including Wikipedia access through natural language queries. While not a library system per se, Wikipedia is one of the most-used educational reference sources.
Notable Library Gaps
Koha, Alma (Ex Libris), Sierra (Innovative), OCLC WorldCat — no MCP servers found for any major integrated library system. Academic libraries are core educational infrastructure, and their absence from the MCP ecosystem is a significant gap. Alma alone serves 3,500+ institutions worldwide.
Education Data
anshumax/world_bank_mcp_server | Community
World Bank open data API access including education indicators — enrollment rates, literacy rates, education expenditure, and more across 200+ countries.
llnormll/world-bank-data-mcp | Community
World Bank Data360 access covering 1,000+ economic and social indicators including education metrics.
Notable Education Data Gaps
NCES (National Center for Education Statistics) — the primary U.S. federal source for education data. Zero MCP servers found.
UNESCO Institute for Statistics — the global authority on education data. Zero MCP servers found.
IPEDS (Integrated Postsecondary Education Data System) — comprehensive U.S. college and university data. Zero MCP servers found.
Common Data Set — standardized data published by most U.S. colleges. Zero MCP servers found.
Architecture Patterns
AI Teaching Assistant
┌─────────────────────────────────────────────────────┐
│ AI Teaching Assistant │
│ (Claude, GPT, Gemini, etc.) │
├──────┬──────┬──────┬──────┬──────┬─────────────────┤
│Canvas│Grade-│Zotero│Slack │Open │ xAPI Learning │
│ MCP │scope │ MCP │ MCP │Edu │ Record Store │
│ │ MCP │ │ │ MCP │ MCP │
├──────┼──────┼──────┼──────┼──────┼─────────────────┤
│Course│Grade │Read- │Class │OER │ Student Learning│
│Mgmt │Submit│ing │Comms │Search│ History │
│ │ions │Lists │ │ │ │
└──────┴──────┴──────┴──────┴──────┴─────────────────┘
An AI teaching assistant with MCP connections to Canvas (course management), Gradescope (grading), Zotero (reading lists), Slack (class communication), OpenEdu (open educational resources), and an xAPI-based learning record store (student history). The agent could answer student questions with course-specific context, draft grading rubrics, suggest supplementary readings, send assignment reminders, and adapt responses based on what topics a student has already mastered.
Research Literature Pipeline
┌───────────────────────────────────────────────────┐
│ Research AI Agent │
├──────┬───────┬──────┬──────┬──────┬──────────────┤
│arXiv │Sem. │Cross │Open │PubMed│ Zotero │
│ MCP │Scholar│Ref │Alex │ MCP │ MCP │
│ │ MCP │ MCP │ MCP │ │ │
├──────┴───────┴──────┴──────┴──────┼──────────────┤
│ Search + Discovery │ Organize + │
│ (find relevant papers) │ Cite │
└───────────────────────────────────┴──────────────┘
│ │
▼ ▼
Literature Review BibTeX Export
(synthesized from (formatted for
multiple databases) LaTeX/Word)
A research agent that searches across arXiv, Semantic Scholar, CrossRef, OpenAlex, and PubMed simultaneously, then organizes findings in Zotero with proper citations. This pipeline can produce comprehensive literature reviews in hours rather than weeks, searching databases that a single researcher might not think to check.
Automated Grading Workflow
┌─────────────────────────────────────────┐
│ AI Grading Assistant │
├──────┬──────┬───────────┬──────────────┤
│Canvas│Grade-│ Winston │ Anki │
│ MCP │scope │ AI MCP │ MCP │
│ │ MCP │(plagiarism│ │
│ │ │ detect) │ │
├──────┼──────┼───────────┼──────────────┤
│Fetch │Apply │ Check │ Generate │
│Sub- │Rubric│ Academic │ Review │
│mis- │Score │ Integrity │ Flashcards │
│sions │ │ │ from Errors │
└──────┴──────┴───────────┴──────────────┘
│
▼
Feedback Report
(per student, with
study recommendations)
An AI grading assistant that fetches submissions from Canvas, applies rubric-based scoring through Gradescope, checks academic integrity with Winston AI, and generates personalized Anki flashcards based on common errors — turning the grading process into a learning feedback loop. The agent produces detailed feedback reports with specific study recommendations for each student.
Accessible Learning Environment
┌─────────────────────────────────────────────────┐
│ Accessibility AI Agent │
├──────┬──────┬──────┬──────┬────────────────────┤
│Canvas│WCAG │ TTS │Open │ Student MCP │
│ MCP │ MCP │ MCP │Lib │ (knowledge graph) │
│ │ │ │ MCP │ │
├──────┼──────┼──────┼──────┼────────────────────┤
│Course│Audit │Audio │Alt. │ Learning │
│Cont- │Pages │Vers- │Format│ Preferences │
│ent │ │ions │Books │ │
└──────┴──────┴──────┴──────┴────────────────────┘
│
▼
Accessibility Report
+ Remediated Content
An accessibility-focused agent that audits course content in Canvas for WCAG compliance, generates audio versions of text content via TTS, finds alternative format books through Open Library, and adapts content delivery based on individual learning preferences stored in a student knowledge graph. Particularly valuable for disability services offices managing accommodations across many courses.
Privacy and Compliance
Education data carries some of the strictest privacy protections of any sector. Any MCP deployment in education must account for these regulatory frameworks.
FERPA (Family Educational Rights and Privacy Act)
FERPA governs access to student education records at any institution receiving federal funding — effectively all U.S. schools and colleges. Key implications for MCP:
- AI agents accessing student grades, attendance, disciplinary records, or financial aid data must comply with FERPA’s “legitimate educational interest” standard
- Third-party services (including AI providers) must have agreements defining their role as “school officials”
- Students (or parents, for K-12) have the right to inspect and request correction of education records — this extends to records processed by AI agents
- In March 2025, the Department of Education issued an unprecedented mandate requiring state agencies to certify FERPA compliance
- Research shows 96% of EdTech apps share student data with third parties, likely violating FERPA
COPPA (Children’s Online Privacy Protection Act)
COPPA protects children under 13 online. Major revisions take effect June 23, 2025, with full compliance required by April 22, 2026:
- The default shifted from opt-out to opt-in consent for data collection
- Explicit parental consent is now required before sharing children’s data with third parties
- AI agents processing data from elementary or middle school students must comply
- This is particularly relevant for K-12 MCP deployments involving tools like Google Classroom, ClassDojo, or Remind
Cybersecurity
The education sector is heavily targeted: organizations averaged 4,388 cyberattacks per week in Q2 2025, a 31% year-over-year increase. Third-party vendors are responsible for the majority of breaches. MCP servers connecting to student data systems must implement:
- Authentication and authorization (OAuth 2.0, API keys with least-privilege access)
- Encryption in transit and at rest
- Audit logging for all data access
- Rate limiting to prevent abuse
- Data minimization — accessing only the data needed for a specific task
Institutional AI Governance
A growing challenge: 2025-2026 research indicates 81% of nonprofit (including educational) organizations use AI ad hoc, with only 4% having documented workflows and nearly half having no AI governance policy. Institutions adopting MCP need to determine:
- What educational context to share with AI agents and what must remain private
- How to enforce boundaries — which MCP tools can access which data
- How to audit AI agent actions across connected platforms
- How to communicate AI usage to students and parents transparently
Student Information Systems: The Missing Layer
The most significant gap in the education MCP ecosystem is the complete absence of student information system (SIS) integrations.
PowerSchool — holds approximately 45% of the U.S. K-12 SIS market, serving 45+ million students. Has REST APIs. Zero MCP servers found.
Infinite Campus — approximately 20% K-12 SIS market share. Uses SOAP and REST APIs. Zero MCP servers found.
Clever — rostering middleware that connects 95,000+ U.S. schools to EdTech applications using data standards like OneRoster. Zero MCP servers found.
ClassLink — identity and access management for K-12, serving 25+ million students. Zero MCP servers found.
Ellucian (Banner, Colleague) — the dominant SIS vendor for U.S. higher education. Zero MCP servers found.
This gap matters because SIS platforms are the authoritative source for enrollment, demographics, grades of record, transcripts, and scheduling — the foundational data layer that everything else in education builds on. Without SIS MCP connections, AI agents in education work with incomplete pictures of students.
The gap likely reflects two factors: the sensitivity of SIS data (FERPA compliance is non-negotiable), and the enterprise sales model of SIS vendors (who may see MCP as a security risk rather than a feature). Unified API approaches — similar to what Clever already provides for rostering — may emerge as a bridge.
Other Notable Gaps
Content Platforms
- Coursera — 130+ million learners, extensive API. No MCP server found
- Udemy — 70+ million learners. No MCP server found
- Khan Academy — free education platform used globally. No MCP server found
- edX / 2U — MOOC platform. No MCP server found (despite Open edX being open source)
K-12 Specific
- Seesaw — student portfolio platform popular in elementary education. No MCP server found
- Kahoot! — interactive quiz platform used in millions of classrooms. No MCP server found
- Nearpod — interactive lesson platform. No MCP server found
- Clever — mentioned above, the rostering layer connecting schools to apps. No MCP server found
Assessment
- Turnitin — academic integrity platform at 16,000+ institutions. No MCP server found
- Respondus LockDown Browser — exam security. No MCP server found
- ExamSoft — secure assessment. No MCP server found
Interoperability Standards
- SCORM — the legacy e-learning content standard. No MCP server found
- LTI (Learning Tools Interoperability) — the standard for connecting tools to LMS platforms. No MCP server found
- ORCID — the researcher identity standard (well-documented REST API). No MCP server found
- OneRoster — the student rostering standard. No MCP server found
Getting Started by Role
For University Professors
Start with a Canvas or Moodle MCP server matching your institution’s LMS, combined with Zotero for reference management and arXiv or Semantic Scholar for literature search. This combination lets an AI agent manage your course, curate reading lists, and assist with research from a single interface. Add Gradescope’s MCP for automated grading in large courses. Be mindful of FERPA — ensure your AI usage complies with your institution’s data governance policies and avoid sending identifiable student data to external AI services without proper agreements.
For K-12 Teachers
Google Classroom’s MCP server (especially the 67-tool Google Apps Script version) provides the strongest starting point for K-12. Combine with OpenEdu’s grade-level filtered educational resources for curriculum-aligned content discovery. For formative assessment, the Anki MCP server can generate study materials from lesson content. The biggest limitation for K-12 is the absence of SIS, Remind, and ClassDojo MCP servers — the tools teachers use most for administration and parent communication.
For Academic Researchers
The research MCP ecosystem is the strongest in education. Start with arXiv (2,400 stars, well-maintained) and Zotero (2,100 stars, semantic search) as your foundation. Add Semantic Scholar for citation network analysis, CrossRef for DOI resolution, and OpenAlex for broad scholarly coverage. The Academix aggregator provides unified access to five databases. PubMed’s 40-tool server is essential for biomedical research. This combination can automate much of the literature review process.
For Educational Technologists
Evaluate your institution’s LMS MCP options and accessibility needs. Canvas has the most mature MCP ecosystem (five implementations). Layer in WCAG accessibility servers to audit course content compliance. The xAPI learning record store MCP (learnmcp-xapi) is architecturally important — it could enable adaptive learning across platforms. Watch for official vendor MCP servers from major LMS platforms, which will likely emerge as MCP adoption grows.
For Instructional Designers
Combine LMS MCP servers with course generation tools (mcp-tutor, MCPLessonGenerate) and OpenEdu’s curriculum-aligned resources. The ability to search Open Library for supplementary reading, generate flashcards via Anki, and ensure accessibility compliance via WCAG tools creates a comprehensive content development workflow. For multi-language course development, leverage the multi-source research servers to find resources across language barriers.
For Education Policy Researchers
World Bank MCP servers provide cross-country education data access — enrollment rates, literacy rates, and education spending across 200+ countries. Combine with PubMed for education-health intersection research and the multi-source academic aggregators for policy literature reviews. The biggest gap is NCES and UNESCO data — you’ll need to access those directly until MCP servers are built.
Market Context
The AI-in-education market shows enormous growth with wide-ranging estimates: $7–19 billion in 2025 (the range reflects different definitions of what counts as “AI in education”), growing to $9.58 billion by 2026 and potentially $136 billion by 2035 at a 34.5% compound annual growth rate. The broader EdTech market is projected to reach $348 billion by 2030.
2026 is widely described as a tipping point — AI moves from experimental “nice-to-have” to mandatory baseline for educational platforms. Key trends shaping the MCP landscape:
Interoperability over more tools. Success is increasingly measured by connected ecosystems, not individual applications. Platforms that pull curriculum, assessment, intervention, and professional learning into single coherent experiences are winning. MCP aligns directly with this trend.
Agentic AI in classrooms. AI agents performing tasks like creating assignments, checking submissions, and providing feedback — as demonstrated by Google Classroom’s Gemini CLI MCP integration. This requires the kind of multi-platform connectivity that MCP provides.
Digital credentials and skills-based learning. AI-powered personalized learning pathways matched to workforce needs, with verified digital credentials. The xAPI + MCP combination could track learning across formal and informal contexts.
AI governance becoming non-optional. Institutions determining what educational context to share with AI, what must remain private, how to enforce boundaries. MCP’s structured tool interfaces provide the transparency needed for institutional governance.
The MCP server count in education continues to grow, but unevenly. Academic research tools are mature and well-adopted. Canvas has strong community coverage. Everything else — K-12 tools, SIS platforms, content providers, assessment systems — represents significant opportunity. The institutions and vendors that build MCP bridges first will shape how AI integrates with education for the next decade.