MCP server

Connect Claude, Cursor, VS Code or any other MCP client to these docs, and read them as plain markdown through llms.txt and the raw routes.

What this is

This documentation is also served as an MCP server. Specifically, if you are building on top of CultTweaker or COTL MP Steam, this MCP helps you with that. this MCP does not provide direct documentation to the game's source code, only for the mods that are available in this page.

https://cotlmod.infernodragon.net/mcp

The transport is streamable HTTP. There is no API key, no account and no rate limit to sign up for — it serves the same pages you are reading now.

The server is read-only. It can list and read documentation pages, and that is all. It has no access to the mods, your game files or anything else.

Quick example

Most clients take a name and a URL. In Claude Code that is one line:

claude mcp add --transport http cotl-docs https://cotlmod.infernodragon.net/mcp

For clients configured by file, the entry looks like this:

{
  "mcpServers": {
    "cotl-docs": {
      "type": "http",
      "url": "https://cotlmod.infernodragon.net/mcp"
    }
  }
}

That shape works for Cursor, in ~/.cursor/mcp.json for every project or .cursor/mcp.json inside one. VS Code uses the same fields under servers instead of mcpServers, in .vscode/mcp.json. For Claude's desktop and web apps, add it as a custom connector and paste the same URL.

Every page on this site also carries an Add MCP Server item in the menu beside its Copy page button, which installs the server into Cursor in one click, and a Copy MCP Server URL item for pasting into anything else.

Once it is connected, ask the tool something the docs answer and it will fetch the page itself:

Using the cotl-docs server, what does CultTweaker's IdOf return, and why should I not save the value?

What the server exposes

Two tools, both read-only and cached for an hour:

ToolTakesReturns
list-pagesnothingTitle, path, description and URL for every page
get-pageapathThat page's metadata plus its full markdown

A client normally calls list-pages first to find what exists, then get-page for the pages it needs. Paths are the same ones in your address bar, so /docs/culttweaker/modding-api is a valid argument to get-page.

Without an MCP client

The same content is available as plain files, which is often enough:

RouteWhat you get
/llms.txtEvery page's title, description and link, in one index
/llms-full.txtThe entire documentation as a single file
/raw/docs/<path>.mdAny single page as its original markdown

For example, the modding API page in markdown is at /raw/docs/culttweaker/modding-api.md. The Copy page menu on each page links to the same file, and can copy the page straight to your clipboard or hand it to ChatGPT or Claude.

Pasting a link to /llms-full.txt into a chat is the fastest way to give an assistant the whole reference at once, with no setup at all.

Mods and site by InfernoDragon0

Cult of the Lamb is a trademark of Massive Monster and Devolver Digital. This is an unofficial fan project.