Skip to main content

For AI agents and developers

Honen speaks MCP.

Connect any MCP-aware client and it can search your courses, read your Knowledge Base, and drive Honen workflows — all inside the permissions the signing-in account already has. Prefer to just read? The whole platform is published as markdown.

Streamable HTTPOAuthNo API key
MCP endpointhttps://honen.com/api/mcp/mcp

Paste this into your client's MCP settings. A browser window opens, you sign in to Honen and click Allow, and the tools appear. Nothing is shared until you approve it.

Reading, not calling

If you are an AI agent, read llms-full.txt.

Every capability Honen has is published as plain markdown — easier to digest than this page, and free of the layout you do not need. Two files: an index, and the whole corpus in one request. Prefer the second unless you know exactly which page you want.

/llms-full.txtRecommendedThe index plus all 17 topic documents concatenated into one file, with a numbered table of contents. One fetch, complete picture./llms.txtThe index: four sections linking each topic document separately, if you would rather fetch only the one you need.

One click

Add it to your editor.

These open the client you already have installed and hand it the endpoint. You still review and approve the server before anything runs.

Cursor

Opens Cursor and pre-fills the server for you to approve.

Add to CursorDocs
VS Code

Adds Honen to your user-profile mcp.json via Copilot Chat.

Add to VS CodeDocs
VS Code Insiders

Same install, against the Insiders build.

Add to InsidersDocs
LM Studio

Adds Honen as a remote MCP server in LM Studio.

Add to LM Studio

Everything else

Or paste the URL.

Clients without an install link take the same endpoint. Two commands and one settings screen cover the rest.

Claude CodeDocs

One command, then /mcp in the session to authenticate.

Terminal
claude mcp add --transport http honen https://honen.com/api/mcp/mcp
Claude apps and claude.aiDocs

Settings → Connectors → Add custom connector, then paste the endpoint. Claude walks you through sign-in.

ChatGPT

Settings → Connectors → add a custom connector and paste the endpoint. Availability depends on your plan and on whether your workspace admin has allowed custom connectors.

Windsurf, Zed, or anything else

Drop this into the client's MCP config file. The key name is yours to choose.

JSON
{
  "mcpServers": {
    "honen": {
      "type": "http",
      "url": "https://honen.com/api/mcp/mcp"
    }
  }
}

An agent never gets more access than the person holding it.

The connection authenticates as you, over OAuth, and every tool call is checked against your existing workspace permissions. There is no separate API key to leak, and a connected agent cannot read a course, a learner record, or a Knowledge Base document you could not already open yourself. Workspace admins can see and revoke connections at any time.

Signed in? Set it up from your dashboard