Getting Started

Install the plugin, configure an AI agent, and run your first prompt inside Unreal Engine.

Prerequisites

• -Unreal Engine 5.4 or newer (source or Launcher build).
• -Node.js - Required only for the CLI agents (Claude Code, Codex). The plugin searches PATH, npm global, bun, cargo, volta, fnm, asdf, mise, scoop, and Homebrew locations automatically. The local agents (Ollama, LM Studio) need no Node.js.
• -At least one agent set up: a CLI agent (Claude Code or Codex), or a local model server (Ollama or LM Studio). See Agent Setup below.

Installation

1. 1.Purchase Blueprint Stack from the checkout page.
2. 2.Download the plugin from the Downloads tab in your dashboard.
3. 3.Copy the BlueprintStack folder into your project's Plugins/ directory.
4. 4.Enable the plugin in Edit > Plugins and restart the editor.

Opening the Chat Panel

Once the plugin is enabled, open the chat panel from the menu bar:

The panel opens as a dockable tab. You can dock it alongside your Blueprint editor, Content Browser, or any other panel.

Agent Setup

Blueprint Stack ships with four built-in agents: two cloud coding agents (Claude Code, Codex) and two local options that run open models on your own machine (Local (Ollama), Local (LM Studio)). Pick whichever fits; you can switch agents per session.

Cloud agents (Claude Code, Codex)

Install via npm in your terminal:

These require their own authentication (e.g., running claude login or setting an API key environment variable before first use). Full per-agent steps: Claude Code, Codex.

Local models (Ollama, LM Studio)

These run open models on your own hardware — no API key, no Node.js, no extra adapter, and nothing leaves your machine. There's nothing to install for Blueprint Stack itself; you just need the Ollama or LM Studio app installed, with its server running and a model loaded:

• -Ollama — install from ollama.com, run ollama pull llama3.1, and make sure the Ollama app (or ollama serve) is running. Full steps: Local (Ollama).
• -LM Studio — install from lmstudio.ai, download a model, then open the Developer tab and click Start Server. Full steps: Local (LM Studio).

Pick the model from the dropdown after selecting the agent; choose a model that supports tool / function calling so it can drive the editor. Host and default-model settings live in the Blueprint Stack panel under Settings > Local LLM (Ollama) and Settings > LM Studio.

Custom Agents

You can register additional agents in Project Settings > Plugins > Blueprint Stack > Agents. Any ACP-compatible CLI can be added here.

Agent Picker

The agent picker dropdown shows a status label next to each agent:

• -Available - The CLI and API key are both detected. Ready to use.
• -Not Installed - The CLI binary was not found on any search path.
• -Missing API Key - The CLI is installed but its required API key is not set.

Path Resolution

The plugin automatically searches for agent CLIs and Node.js across the following locations:

• -System PATH
• -npm global install directory
• -bun, cargo, volta, fnm, asdf, mise
• -scoop (Windows), Homebrew (macOS/Linux)

Your First Prompt

1. 1.Select an agent from the dropdown (make sure it shows Available).
2. 2.Type a message in the input field, for example:

3. 3.The agent will call the read_file tool, read the Blueprint's structure, and respond with a summary.

Attaching Context

You can provide additional context to the AI by attaching Blueprint nodes, assets, or images:

• -Ask AI about a node - Right-click a node in the Blueprint graph and select Ask AI about this node. The node's pins, connections, and properties are serialized and sent with your next prompt.
• -Add asset to context - Right-click an asset in the Content Browser and select Add to AI context. The asset's structure (variables, components, functions, events) is serialized.
• -Drag and drop - Drag assets directly from the Content Browser onto the chat input area.
• -Paste images - Paste an image from your clipboard (Ctrl/Cmd+V) into the chat input. The image is base64-encoded and sent to vision-capable models.

What's Next?

• -Explore the full list of editor tools in the Tools Reference.
• -If anything fails during setup, check the Troubleshooting page for quick fixes (Node path issues, missing CLIs, API key configuration, and more).