Ariescar/anyCreature

You can now generate creatures directly within your project. Drop the open-source into your Agent, and start prompting what you want.

What it solves

anyCreature allows users to generate game-ready 3D creatures from simple text descriptions. It eliminates the need for manual 3D modeling, mesh files, or photogrammetry by using an AI-driven pipeline that converts text orders into fully skinned, animated, and vertex-colored GLB files.

How it works

The project uses a multi-stage pipeline (defined by "cards") that an AI agent (like Claude Code or Cursor) executes. The process follows these steps:

  1. Interview: The agent asks a few clarifying questions based on a text order.
  2. Design: The agent creates a silhouette brief and a low-fidelity design.
  3. Quality Gates: The design is passed to a context-free "reader agent" that must be able to recognize the creature from four silhouette views (including a 24px thumbnail). If the design is not recognized or is not "punchy" enough, it is sent back for iteration.
  4. Compilation: A custom engine (ACS engine) compiles a JSON specification into a skinned GLB file.
  5. Finalization: The engine handles vertex coloring, ambient occlusion baking, and animations (idle, move, attack).

Who it’s for

Game developers and "vibe coders" who need high-quality, 3D creature assets quickly without needing professional 3D modeling software or manual asset creation.

Highlights

  • Text-to-3D Pipeline: Converts a one-line text order into a game-ready asset in one session.

  • Context-Free Validation: Uses separate AI agents to act as blind readers to ensure the creature's silhouette is recognizable and bold.

  • Zero-Dependency Engine: The core compiler (ACS engine) requires only Node.js to turn a JSON spec into a GLB.

  • Game-Ready Output: Produces GLB files that are skinned, animated, and vertex-colored with no external textures.

  • Grafting Capabilities: Every GLB embeds its own source spec, allowing agents to extract and graft parts between different creatures.

Related

  • Project
    OpenDCAI/GameFactory-3A

    An open-source framework that uses coding agents to transform game requirements into production-ready assets and engine-specific code for UE5, Unity, Godot, Blender, and three.js.

  • Project
    PhiloLabs/fable51-worlds

    A system that uses AI agent swarms to generate verifiable, walkable 3D worlds from text, images, or video, rendered as code-based Three.js applications.

  • Project
    HorizonRobotics/EmbodiedGen

    A generative 3D world engine that converts text and images into simulation-ready 3D assets and environments for embodied AI, deployable across multiple robotics simulators.

  • Project
    HKUDS/CLI-Anything

    CLI‑Anything is a Python framework that automatically generates agent‑friendly Click CLIs for any software, outputting JSON for easy parsing. It includes a 7‑phase generation pipeline (analyze → design → implement → test → document → publish) and a package manager called CLI‑Hub to browse, install, and run a growing catalog of pre‑built CLIs (e.g., GIMP, Blender, Zoom). The project targets AI agents (Claude Code, Cursor, OpenClaw, etc.) that need deterministic, composable command interfaces, and it ships with a comprehensive test suite and security hardenings.

  • Project
    RareSense/Nova3D

    Nova3D is a code-native 3D generation system that creates assets as executable Blender Python scripts, enabling the production of structured, editable, and animatable 3D models with named parts.