Use Gameheim with your AI assistant

What you can ask

A few examples from producers, designers, QA and programmers. Write them in your own words: the assistant picks the right Gameheim tool.

• Which blocker bugs are still open in OFT, and who is working on them?
• Here are my notes from today’s meeting: create a task for each action item and assign it.
• Report this crash as a bug, with the log I am pasting, and create the task that fixes it.
• What did we decide about the inventory system, and why?
• What is left for me this week? Move the task I just finished to done.
• Search the wiki: how do we name level streaming volumes?

Before you start

• Your role needs the External apps (MCP & API) permission. Your LEAD grants it in the role editor; read-only accounts cannot connect apps.
• An assistant can do only what you can do, and only in the projects you choose when you connect.
• Nothing can be deleted through Gameheim’s tools.

The address of Gameheim for your assistant is:

https://api.gameheim.com/api/mcp

Connect your assistant

Menus change from one release to the next: if a step does not match what you see, follow the assistant’s own guide linked in each box.

claude mcp add --transport http gameheim https://api.gameheim.com/api/mcp

{
  "mcpServers": {
    "gameheim": {
      "url": "https://api.gameheim.com/api/mcp"
    }
  }
}

{
  "servers": {
    "gameheim": {
      "type": "http",
      "url": "https://api.gameheim.com/api/mcp"
    }
  }
}

What you approve

The first time, your assistant opens Gameheim in the browser. Sign in, check the name of the app and where it will send you back, choose what it may do and in which projects, then confirm with your password (and your verification code, if two-step verification is on).

You can disconnect an app at any time in Integrations, AI assistants. Your LEAD sees the apps connected in the organization and can disconnect them too. A connection closes by itself after 90 days without use and after one year in any case.

Assistants ask first for read access only. When one needs to write, it asks you to approve the extra permission.

Your data

What a tool returns becomes part of your conversation with the assistant, and so reaches its provider (for example Anthropic or OpenAI) under your agreement with them. That is why the permission is granted by your LEAD, and why you choose the projects each app can reach.

Every change an assistant makes is recorded in the audit log with the app that made it.

Tools

For the technically curious: Gameheim connects through the Model Context Protocol (MCP), and these are the tools your assistant can call. Tasks, bugs and decisions have a key such as OFT-12, OFT-BUG-5 or OFT-DEC-3. Texts are Markdown.

Workflows for your team

Bug triage, meeting notes turned into tasks, decision records and project knowledge search, role by role: see AI assistants for game projects.

Building your own integration?

The same operations are available as a REST API, in preview. See the API reference.