The Examplary MCP Server provides AI assistants with comprehensive access to Examplary's exam management platform through 60+ API endpoints.
Core capabilities include:
Exam Management - Create, read, update, delete, duplicate, export (PDF/Word), and import exams from various formats (Word, PDF, plain text, Moodle XML). Organize exams using folders and track metadata.
AI-Powered Question Generation - Generate new questions and update existing ones using AI prompts and source materials. Get context suggestions for exam generation and cancel ongoing generation jobs.
Question Operations - Browse public and custom question types, create organization-specific types, enable/disable types for workspaces, and manage grading instructions and generation parameters.
Question Bank - Store, manage, and organize reusable questions with tags for easy retrieval.
Source Materials - Upload and process documents (PDF, Word, etc.) with AI to extract facts, topics, and summaries. Create page slices from materials and track processing status.
Student Assessment & Grading - Create and manage student sessions, scan answers from uploaded documents, provide AI-powered automatic grading with suggestions, accept/reject grading suggestions individually or in bulk, and generate overall feedback.
Workspace & Collaboration - Create and manage organization workspaces, invite team members with role-based access, update workspace settings, and browse publisher libraries for public content.
Taxonomies & Standards - Access default taxonomies (e.g., Bloom's Taxonomy), create custom taxonomies for learning objectives, and define taxonomy levels with AI generation instructions.
User & Media Management - View and update user profiles, manage preferences and workspace memberships, generate signed upload URLs, and handle media files securely.
Additional Features - Import questions from multiple file formats, browse student levels for age-appropriate content, and filter by external IDs for system integration.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Examplary MCP Server create a new exam about machine learning fundamentals"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A Model Context Protocol (MCP) server for Examplary, providing AI assistants with access to the Examplary exam management API.
60+ API Endpoints - Full access to Examplary's exam management capabilities
Auto-generated Tools - All API operations automatically available as MCP tools via OpenAPI specification
Secure Authentication - User-specific API key authentication
Cross-platform - Works on macOS, Linux, and Windows via UV runtime
One-click Installation - Install as a .mcpb package in Claude Desktop
Related MCP server: MCPaeroedu
Examplary is an AI-powered exam management platform that helps educators:
Create and manage exams with AI assistance
Generate questions from source materials
Grade student responses automatically
Organize content in collaborative workspaces
Share resources via publisher libraries
Download the latest examplary-mcp.mcpb from the releases page
Open Claude Desktop
Go to Settings → Extensions
Click "Install Extension" and select the downloaded .mcpb file
When prompted, enter your Examplary API key
Clone this repository:
git clone https://github.com/examplary-ai/mcp.git
cd mcpInstall UV (if not already installed):
curl -LsSf https://astral.sh/uv/install.sh | shRun the server:
export EXAMPLARY_API_KEY="your-api-key-here"
uv run src/server.pyLog in to Examplary
Navigate to Account → Developer
Click "Generate New API Key"
Copy the API key and save it securely
Note: Keep your API key secret. Never commit it to version control or share it publicly.
Once installed in Claude Desktop, you can ask Claude to interact with Examplary:
"Create a new exam about Python programming"
"List all my exams"
"Generate questions from this document"
"Get the results for exam ID 12345"
"Create a new organization workspace"
The MCP server provides access to all Examplary API endpoints, including:
Create, read, update, and delete exams
Duplicate exams
Export to PDF or Word
AI-powered question generation
Store and organize reusable questions
Public and private question types
Bulk import from various formats
Create grading sessions
Scan documents for answers
AI-powered grading and feedback
Accept or reject grading suggestions
Manage workspaces
Invite team members
Organize exams in folders
Browse public content
Share exam templates
Discover featured resources
examplary-mcp/
├── manifest.json # MCPB manifest configuration
├── pyproject.toml # Python dependencies
├── src/
│ ├── __init__.py
│ └── server.py # Main MCP server
├── .mcpbignore # Files excluded from .mcpb package
└── .github/
└── workflows/
└── release.yml # Automated release workflowInstall dependencies:
uv pip install -e .Set your API key:
export EXAMPLARY_API_KEY="your-api-key"Run the server:
uv run src/server.pyTo build the .mcpb package manually:
zip -r examplary-mcp.mcpb \
manifest.json \
pyproject.toml \
src/ \
icon.png \
-x "*.pyc" -x "__pycache__/*" -x ".git/*"The server uses stdio transport for communication with Claude Desktop. To test locally:
Run the server with your API key set
The server will start and listen for MCP protocol messages on stdin
Send MCP requests via stdin and receive responses on stdout
Examplary enforces rate limits on certain operations:
Exam generation: 3 requests per 60 seconds
Question operations: 10 requests per 60 seconds
Other endpoints: Standard rate limiting applies
API keys are stored securely by Claude Desktop
All communication uses HTTPS
API keys can be revoked at any time from the Examplary dashboard
Never share your API key or commit it to version control
Examplary Documentation: https://developers.examplary.ai
Examplary Support: support@examplary.ai
MIT License - see LICENSE file for details
Contributions are welcome! Please feel free to submit a Pull Request.
Fork the repository
Create your feature branch (git checkout -b feature/amazing-feature)
Commit your changes (git commit -m 'Add some amazing feature')
Push to the branch (git push origin feature/amazing-feature)
Open a Pull Request
Built with:
FastMCP - Python framework for MCP servers
Model Context Protocol - Open protocol for AI-application integration
Examplary API - Exam management platform
Create a new custom attribute in the current workspace.
Inputs: name, attributeType, options, showInStudentDashboard, +3 more.
Delete a custom attribute from the current workspace. Default attributes cannot be deleted.
Inputs: id (required).
Get all custom attributes configured for the current workspace.
Inputs: no named parameters.
Update the position of multiple attributes at once.
Inputs: order.
Update an existing attribute in the current workspace.
Inputs: id (required), name, attributeType, options, +4 more.
Delete a rubric from the current workspace.
Inputs: id (required).
Create a new embed session. This allows you to embed the exam generation flow into your own application.
Inputs: body. Returns: id, status, embedUrl, flow, +9 more
Retrieve an embed session by its ID.
Inputs: id (required). Returns: id, status, embedUrl, flow, +9 more
Revoke access to an embed session by its ID.
Inputs: id (required).
Cancel an ongoing job to generate new questions for the exam using AI.
Inputs: id (required).
Create a new exam within your workspace.
Inputs: name, sourceMaterialIds, context, studentLevel, +8 more.
Delete the exam.
Inputs: id (required).
Duplicate the exam's questions and settings to a new exam.
Inputs: id (required), name, includeSessions.
Export an exam as a QTI 2.1 ZIP package.
Inputs: id (required). Returns: url
Export an exam as a QTI 3 ZIP package.
Inputs: id (required). Returns: url
Get a single exam by its ID.
Inputs: id (required).
Get AI-generated suggestions for context/instructions to include in the exam generation prompt, based on the source materials linked to the exam.
Inputs: id (required).
Create an exam by importing one or more source materials (e.g. an existing exam).
Inputs: name, sourceMaterialIds.
Get a list of all exams within your workspace.
Inputs: folder, role.
Export an exam to a PDF file or Word document.
Inputs: id (required).
Generate a new question for the exam using an AI prompt.
Inputs: id (required), message, questionType.
Retrieve a question by its ID.
Inputs: id (required), questionId (required), format.
Import questions into an exam from a file. Support Word, PDF, Moodle quiz XML and plain text files.
Inputs: id (required), name, type, url, +2 more.
Update a question using an AI prompt.
Inputs: id (required), questionId (required), message.
Accept all AI-generated grading suggestions for the exam session.
Inputs: id (required), sessionId (required).
Accept a single AI-generated grading suggestion for the exam session.
Inputs: id (required), sessionId (required), questionId (required).
Remove an exam session.
Inputs: id (required), sessionId (required).
Edit a single answer in an exam session.
Inputs: id (required), sessionId (required), questionId (required), answer, +1 more.
Generate feedback for the test as a whole in an exam session.
Inputs: id (required), sessionId (required).
Get a single exam session by its ID, representing a set of answers from a student.
Inputs: id (required), sessionId (required).
Manually create a session, possibly from uploaded data.
Inputs: id (required), studentName, source, answers.
Get a list of all student sessions for an exam, representing a set of answers from a student.
Inputs: id (required).
Set teacher feedback and/or a grade for a specific question in an exam session.
Inputs: id (required), sessionId (required), question, pointsAwarded, +1 more.
Set teacher feedback for the test as a whole in an exam session.
Inputs: id (required), sessionId (required), feedback.
Scan student answers from a document.
Inputs: id (required).
Start a background task to extract student answers from an uploaded PDF document.
Inputs: id (required), documentUrl.
Initiate a job to generate new questions for the exam using AI, based on the source materials and metadata specified.
Inputs: id (required).
Update an existing exam.
Inputs: id (required).
Create a new folder in the current workspace.
Inputs: name.
Delete a folder from the current workspace. If there are any exams in the folder, they will be moved out of the folder.
Inputs: id (required).
Get a list of folders for exam organisation that exist within the current workspace.
Inputs: no named parameters.
Update an existing folder in the current workspace.
Inputs: id (required), name.
Get a list of rubrics that exist within the current workspace.
Inputs: no named parameters.
Add a member to a group. Requires group manager/owner role or org admin/owner.
Inputs: id (required), body.
Create a new group. Requires admin or owner org role.
Inputs: name.
Delete a group and revoke all associated permissions. Requires group manager/owner role or org admin/owner.
Inputs: id (required).
List all groups in the organization. When restricted group visibility is enabled, only returns groups the user belongs to.
Inputs: no named parameters.
List all members of a group with resolved user info. Respects restricted group visibility.
Inputs: id (required).
Remove a member from the group. Requires group manager/owner role or org admin/owner. Cannot remove the owner.
Inputs: id (required), userId (required).
Rename a group. Requires group manager/owner role or org admin/owner.
Inputs: id (required), name.
Change a member's role in the group. Requires group manager/owner role or org admin/owner.
Inputs: id (required), userId (required), role.
Cancel a background job.
Inputs: id (required).
Poll the status of a background job.
Inputs: id (required).
Returns a signed URL for uploading a file, and a public URL for accessing it afterwards.
Inputs: filename, type, contentType. Returns: uploadUrl, publicUrl
Get the current user's account details.
Inputs: no named parameters.
Update the current user's account details.
Inputs: name, preferences.
Get custom domain configuration
Inputs: no named parameters.
Update custom domain configuration
Inputs: customDomain.
Delete organization
Inputs: no named parameters.
Get organization
Inputs: no named parameters.
Create a new workspace to collaborate with others on exams.
Inputs: name.
List all organizations
Inputs: no named parameters.
Update organization
Inputs: name, logoImageUrl, menuColor, accentColor, +2 more.
Update an existing rubric in the current workspace.
Inputs: id (required), name, scoring.
Create a new permission for a specific resource, or update the role of an existing actor.
Inputs: resource (required), body. Returns: actor, role, createdAt, updatedAt, +1 more
Search for users and groups to share with. Returns matches by name or email. Respects restricted group visibility.
Inputs: q (required), type.
Remove a permission for a specific actor on a specific resource.
Inputs: resource (required), actor (required).
Get the role of a specific actor for a specific resource.
Inputs: actor (required), resource (required). Returns: actor, role, createdAt, updatedAt, +1 more
Get a list of users, groups and orgs that have access to a specific resource.
Inputs: resource (required). Returns: result
Add a rubric to the current workspace.
Inputs: name, scoring.
Start a background job to generate a rubric using AI.
Inputs: message, sourceMaterialIds, language, context.
Cancel an in-progress mastery topic generation job.
Inputs: id (required).
Create a new practice space within your workspace.
Inputs: name, sourceMaterialIds, masteryTopics, settings, +5 more.
Delete the practice space.
Inputs: id (required).
Duplicate the practice space's settings and topics to a new practice space.
Inputs: id (required), name.
Get feedback for a specific topic in a practice space.
Inputs: id (required), topicId, feedbackType.
Start generating mastery topics for a practice space.
Inputs: id (required).
Get a single practice space by its ID.
Inputs: id (required).
Get the progress of a single practice space by its ID.
Inputs: id (required).
Get a list of all practice spaces you have access to in this workspace.
Inputs: folder, role.
Get a preview of the top practice space questions by quality score.
Inputs: id (required).
Create a new practice space session.
Inputs: id (required), prompt.
Get a practice space session by its ID.
Inputs: id (required), sessionId (required).
Get the current user's past practice space session.
Inputs: id (required).
Save an answer for a specific question in a practice space session.
Inputs: id (required), sessionId (required), questionId (required), question, +3 more.
Get details about a specific practice space student.
Inputs: id (required), studentId (required).
Get a list of all students that practiced in a practice space.
Inputs: id (required).
Update an existing practice space.
Inputs: id (required), name, settings, folder, +5 more.
Add a question to the question bank in the current workspace.
Inputs: question.
Delete a question bank item from the current workspace.
Inputs: id (required).
Get a list of question bank items that exist within the current workspace.
Inputs: no named parameters.
Update an existing question bank item in the current workspace.
Inputs: id (required), question.
Deletes a specific question type by its ID. Note that only the owner organization of the question type can delete it, and only if it has not been used in an exam.
Inputs: id (required).
Disable a question type for the current workspace.
Inputs: id (required).
Enable a question type for the current workspace.
Inputs: id (required).
Retrieves a specific question type by its ID.
Inputs: id (required).
Get JS module for a question type PCI for QTI 3 export.
Inputs: id (required).
Lists all question types, either those available by default or those created within the user's organization.
Inputs: no named parameters.
Lists all public question types, which can be enabled in workspaces.
Inputs: no named parameters.
Creates a new question type or updates an existing one. If the question type already exists, it will be updated; otherwise, a new one will be created.
Inputs: $schema, id, name, description, +26 more.
Add a source material and start processing it for later use in an exam.
Inputs: url, type, name, size, +3 more. Returns: org, id, type, createdBy, +16 more
Deletes the specified source material.
Inputs: id (required).
Returns the current status and facts extracted from the specified source material.
Inputs: id (required). Returns: org, id, type, createdBy, +16 more
Returns a list of source materials for the current organization.
Inputs: externalId. Returns: result
Create a new source material based on a specific page range of an existing source material.
Inputs: id (required), startPage, endPage, name. Returns: org, id, type, createdBy, +16 more
Updates the specified source material.
Inputs: id (required), name.
Get a list of available default student levels.
Inputs: no named parameters.
Save a new taxonomy in the current workspace.
Inputs: name, description, instructions, levels.
Delete a taxonomy from the current workspace. This fails if there are any exams associated with the taxonomy.
Inputs: id (required).
Get a list available taxonomies, including defaults and taxonomies created in the current workspace.
Inputs: no named parameters.
Update an existing taxonomy in the current workspace.
Inputs: id (required), name, description, instructions, +1 more.
Create a new user in the workspace.
Inputs: email, name, role, sendInvite.
Delete a user from the workspace.
Inputs: id (required).
Get a list of all users in the workspace.
Inputs: no named parameters.
Update a user's role within the workspace.
Inputs: id (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 115 tools
Most tools are grouped by resource (e.g., exams_, groups_), making their purpose clear. However, some tool names like exams_import and sourceMaterials_create could be confused without descriptions, and the mix of singular/plural resource names (e.g., questionBank vs studentLevels) causes minor ambiguity.
Naming conventions are inconsistent: some use snake_case (attributes_create), some use camelCase for the resource prefix (embedSessions_create), and a few use no separator (deleteRubricsid). This mix of styles makes the set feel unpolished.
At 115 tools, the count is very high for an MCP server, likely overwhelming for agents. While the domain (exam management) is broad, many tools could be consolidated (e.g., separate tools for each exam session operation).
The tool surface covers the full lifecycle of exams, sessions, groups, rubrics, question banks, and more, including AI generation and export. Minor gaps exist (e.g., no bulk delete or analytics), but core workflows are well-supported.
Connect AI assistants to Edusign attendance, courses, learners, documents, and API docs via OAuth.
Connect AI assistants to Stellary projects, boards, documents, and governed agent workflows.
Connect AI agents to 580+ APIs with OAuth or API-key auth. Hosted gateway; API pricing varies.
Find and call verified APIs from any AI assistant, with free trial calls and previewed writes.
