Core Features
Dynamic Editor Environment
Customizable Templates
Resilient JSON Parsing
Built natively on Unity’s modern UI Toolkit architecture. The two-pane split view pairs a smooth chat dialogue interface with a real-time, live design document preview editor.
Driven completely by a modular JSON structure. You can seamlessly customize document sections, descriptions, and examples to perfectly match your team's specific workflow.
Equipped with an advanced, handcrafted parsing engine that features backslash parity validation and precise pointer tracking. It guarantees immune-level stability against unstable LLM outputs.
SOVEREIGNTY
Architecture & Security
Data Security
DevForge AI prioritizes the absolute security of your intellectual property. By supporting localized inference engines like Ollama out of the box, the tool ensures that your unreleased mechanics, core loops, and studio secrets remain strictly within your local machine or corporate intranet.
Risk Control
This architecture entirely eliminates the geopolitical and compliance risks associated with transmitting raw ideas to foreign cloud servers or unverified third-party APIs.
Full Ownership
You retain 100% data sovereignty while building your next masterpiece.
Get Started in 3 Steps
Step 1: Setup Local LLM
Run your choice of local model (e.g., llama3) via Ollama or specify your preferred commercial cloud endpoint.
Step 2: Open AI Architect Window
Navigate to Tools -> AI Game Design Architect inside Unity. Configure your Endpoint URL and Model Name in the Project Settings.
Step 3: Start Brainstorming
Pitch your concept freely. The AI will intelligently interview you, flesh out the Markdown sections, and let you export a pristine .md specification file when finished.
Support Contact
Need assistance? For bug reports, template customization inquiries, or feature requests, please use the
Contact us here
form above. Our technical team will respond to your inquiry via email as soon as possible.
Privacy Policy Statement
Privacy & Governance: We do not collect, track, or upload your prompt history or document contents. All network traffic consists of direct API requests originating from your local editor environment to your user-configured endpoints.
