AI & Agents
Thought Industries MCP Server
A standards-based Model Context Protocol server that exposes your learning platform as tools any MCP-compatible AI client can use — Claude, IDE assistants, and custom agents.
Launching soon — The MCP Server is not yet available. This page documents the planned functionality.
Overview
The Model Context Protocol (MCP) is an open standard for connecting AI assistants to external tools and data sources. The Thought Industries MCP server exposes your platform's learners, courses, and content as tools that any MCP-compatible client can invoke — including Claude, Cursor, and other AI development tools.
Endpoint
The MCP server is available via Streamable HTTP at:
https://your-instance.thoughtindustries.com/ai/mcpReplace your-instance with your platform subdomain.
Authentication
Authenticate with a Bearer token in the Authorization header. All requests must include the following headers:
| Header | Value | Description |
|---|---|---|
Authorization | Bearer <YOUR_TI_API_KEY> | Scoped API key generated from your Thought Industries admin panel. |
Content-Type | application/json | Request body encoding. |
Accept | application/json, text/event-stream | Required. Requests missing this header are rejected with 406 Not Acceptable. |
Connect a client
Add the Thought Industries MCP server to your client's configuration:
{
"mcpServers": {
"thought-industries": {
"url": "https://your-instance.thoughtindustries.com/ai/mcp",
"headers": {
"Authorization": "Bearer <YOUR_TI_API_KEY>"
}
}
}
}Verify the connection with a cURL request:
curl -X POST \
https://your-instance.thoughtindustries.com/ai/mcp \
-H "Authorization: Bearer <YOUR_TI_API_KEY>" \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-d '{"jsonrpc":"2.0","method":"tools/list","id":1}'Tools & capabilities
The MCP server exposes the following tools. Tools that modify data require write permissions on your API key.
| Tool | Description |
|---|---|
list_courses | Returns a paginated list of courses available on the platform. Supports filtering by status, category, and search term. |
get_learner | Retrieves profile and enrollment details for a specific learner by ID or email address. |
list_enrollments | Lists enrollments for a given learner or course. Includes completion status and progress percentage. |
enroll_user | Enrolls a learner in a specified course. Requires write permissions on the API key. |
create_course | Creates a new course with the provided title, description, and configuration. Requires write permissions. |
Write permissions: enroll_user and create_course require a key with write scopes. Read-only keys will receive a permission error when calling these tools.
Security
| Practice | Detail |
|---|---|
| HTTPS only | The MCP endpoint is only accessible over HTTPS. Plaintext HTTP connections are rejected. |
| Scope your API key | Generate a dedicated key with only the permissions your AI workflow needs. Avoid admin-level keys. |
| Rotate keys regularly | Rotate your API key on a regular schedule and immediately if you suspect compromise. |
Requests missing the Accept header will be rejected with 406 Not Acceptable. Ensure your client sends Accept: application/json, text/event-stream with every request.