PromptBlueprint: LLM Integration for Unreal Engine

PromptBlueprint is an Unreal Engine plugin that enables seamless communication with Large Language Models (LLMs) directly from Blueprints and C++.

It allows you to send prompts to any LLM API endpoint (such as OpenRouter, OpenAI-compatible APIs, or custom servers) and receive responses asynchronously via callbacks.

This makes it ideal for building AI-driven gameplay systems, NPC dialogue, developer tools, procedural generation, and in-editor virtual assistants—without writing complex networking code.

---

✨ Features

• ✅ Blueprint & C++ Accessible: Communicate with LLMs directly from your game logic.
• ✅ Universal API Support: Send prompts to any LLM URL (OpenAI, OpenRouter, custom servers).
• ✅ Asynchronous Requests: Non-blocking HTTP requests keep your game running smoothly.
• ✅ Simple Callback System: Clear success and failure handling for every API call.
• ✅ JSON Field Extraction: Easily parse responses using ordered key paths (e.g., choices[0].message.content).
• ✅ Versatile Use: Ideal for AI chat, tool-calling, gameplay logic, and automation.
• ✅ Lightweight & Modular: Built on native Unreal Engine components for performance.

---

🛠 Supported Use Cases

• AI-Powered NPC Dialogue Systems (dynamic, context-aware conversations)
• Developer Assistant Tools inside Unreal Editor (e.g., generating code snippets, summarizing documentation)
• Procedural Generation of story beats, quests, or dialogue trees at runtime.
• QA Bots and Debugging Assistants for in-editor testing.
• Runtime Prompt-Based AI Logic for emergent gameplay.

---

⚙️ Installation

1. Clone or download this repository.
2. Place the plugin inside your Unreal Engine Project's Plugins directory:

   <YourProjectName>/Plugins/PromptBlueprint/
   

3. Ensure the plugin is enabled in your Unreal Editor's Plugins window.
4. If building from source, you may need to add "PromptBlueprint" to the PublicDependencyModuleNames in your project's Build.cs file.

---

🚀 Getting Started (Blueprint Example)

The core functionality is exposed through a single, easy-to-use Blueprint node: Send LLM Request.

1. Configure the Request

On any Blueprint graph (e.g., in a Widget, Actor, or Blueprint Function Library):

• Target URL: The API endpoint for your LLM (e.g., https://api.openai.com/v1/chat/completions).
• API Key: Your authorization token (often passed in the Authorization header).
• Model Name: The specific model you want to use (e.g., gpt-4o, llama3-70b).
• Prompt/Messages: The text input for the LLM.

2. Connect the Callbacks

The node executes asynchronously and will branch to one of two pins once the network request is complete:

Pin	Description
On Success	Fires when a response is received. Provides the raw Response JSON and the extracted Content String.
On Failure	Fires on network errors, invalid URLs, or HTTP error codes. Provides the Error Message.

3. Extracting the Content

The Send LLM Request node automatically handles extracting the useful response content.

• Extraction Path: Use a key path like choices[0].message.content (standard for OpenAI-compatible APIs) to directly retrieve the final generated text into a simple Content String variable. This avoids manual JSON parsing for the most common use case.

---

💡 C++ Integration

For C++ users, the functionality is exposed via the UPromptBlueprintFunctionLibrary class.

// Example C++ call
FString URL = TEXT("Your API URL");
FString Key = TEXT("Your API Key");
FString Payload = TEXT("{\"model\": \"gpt-3.5-turbo\", \"messages\": [{\"role\": \"user\", \"content\": \"What is Unreal Engine?\"}]}");
FString ContentPath = TEXT("choices[0].message.content");

// Function signature (simplified)
UPromptBlueprintFunctionLibrary::SendLLMRequest(
    URL,
    Key,
    Payload,
    ContentPath,
    // Success/Failure handling via delegates in C++
    ...
);