Try it here https://gg.kiberonlabs.com/app/ A fast, framework-agnostic graph rewriting / graph grammar engine for TypeScript and Rust You can easily embed it into game engines, procedural generation or analytics Projects. The system implements VF2++ as the underlying matching system designed for fast incremental matching, which makes it easy to use the React playground to explore and visually prototype.

I'd always wanted a graph grammar system ever since I read about them. Immediately I knew multiple use cases that I could apply it to and yet it always felt like it was something that has always been primarily within the research domain and didn't have an easy intuitive to use system for people who wanted to explore with it. I focused a lot on making it intuitive and flexible whilst having a powerful playground to interact with during visual authoring. You can either use the typescript package or the Rust crate/DLL to easily embed it into existing systems. Additionally since I had always planned to use this system for procedural quest generation, there is a small questing module made up of my personal insights into questing systems used in games as well as narrative systems like Propps morphology of folklore that I analyzed to try and make more intriguing quests using a simulationist approach. I hope you have as much fun using it as I had making it.

Played around with the React playground for a bit and the visual graph matching actually made it click for me. VF2++ feels snappy even on the messy toy examples I threw at it.

Tried the playground for a few minutes and was surprised how snappy the live graph rewriting felt, especially when I kept editing patterns on the fly. Embedding that into a procedural gen project sounds genuinely useful.

Played around with the React playground for a bit and the visual graph matching actually made it click for me. VF2++ feels snappy even on the messy toy examples I threw at it.

Kiberon Lab | Graph Grammar

Identify and generate content using relationships rules.

Identify and generate content using relationships rules.