Skip to content
Raid Haven modding docs

Using an AI agent

You can use a coding agent such as Claude Code, Codex or Cursor to help with a mod. Give it the plain-text copies of this site. They have the same content as the pages without the menus around them.

File What it holds
llms.txt A short index that points to the two files below
llms-full.txt Every page in one Markdown file
llms-small.txt The same, shortened for agents with less room

If your agent can fetch web pages, give it the address of llms-full.txt. If it can’t, save the file into your mod’s folder and tell the agent to read it.

On a wide screen, most pages also have Copy page as Markdown under “On this page”, for pasting a single page into a chat.

Open your mod’s folder in the agent, paste this, then write your idea after the last line.

Make a mod
I'm making a mod for Raid Haven, a game by Tabbing Studios. This folder is the mod.

First read https://modding.tabbingstudios.com/raid-haven/llms-full.txt. It is the game's modding documentation in one Markdown file.

Follow the docs exactly. Only use file names, folder layouts, field names and functions that the docs show, and copy their examples where you can. If my idea needs something the docs don't cover, stop and tell me instead of guessing.

Before you write anything, tell me which parts of my idea are data and which need a Lua script, and which pages you are following.

When you're done, list the files you made and tell me what to check in the Mod Lab.

This is my idea.

The Mod Kit in the ModKit folder next to the game has an AGENTS.md, a file written for coding agents. Copy it into your mod’s folder and most agents read it on their own. It also has the Lua annotation files that give autocomplete in VS Code, for content fields and for the scripting API. An agent can read those too.

An agent follows the same rules you do, and it can still get things wrong. Load what it made in the Mod Lab and read the Problems tab before you publish it.