Single-spec creature compilation
The engine compiles one JSON specification into a skinned GLB, so the creature is described in text rather than assembled from mesh files or external art packs.
anyCreature
anyCreature is a GitHub project that turns a short text order into a game-ready 3D creature. It compiles a single JSON spec into a skinned GLB with animation, vertex colour, ambient occlusion baking, and an offline showroom viewer.
What anyCreature does
anyCreature is a GitHub-hosted project for turning a short text order into a game-ready 3D creature in a single session. The repository describes an AI-assisted workflow that can ask up to two questions, then compile one JSON spec into a skinned GLB with animation, vertex colour, ambient occlusion baking, and an offline showroom viewer.
The project is positioned as a text-to-creature pipeline rather than a mesh authoring tool. The README emphasizes that it does not rely on downloaded art packs or photogrammetry, and that the engine interface is intentionally small: `node engine/cli.js spec.json out.glb`. The documentation also includes a manual, stage cards, calibration assets, and harness tools for measuring silhouettes and enforcing output checks.
The worked example in the repository is `example/wolf.json`, described as a light quadruped built from plain text and output by the engine. The repo notes that the example ships with two animations and that some features, such as `build: "rigid"`, are intended for non-organic subjects rather than animals.
Core capabilities
Structured multi-stage generation
The stated workflow supports an order, up to two interview questions, silhouette design passes, and a final ship stage that stamps and delivers the result.
Turnkey deliverable package
The output includes animation, vertex colour, ambient occlusion baking, and an offline showroom viewer, according to the repository README.
Built-in calibration checks
The repository includes calibration content that must pass and be blocked in the right cases, so the checker can distinguish acceptable and unacceptable outputs on a given machine.
Inspectable generation controls
The README exposes controllable spec options such as `smooth_angle`, `shading`, `build: "rigid"`, and `ao`, which let authors tune silhouette softness, colour treatment, rigidity, and AO baking.
Role-based quality presets
The docs call out preset QC roles under `harness/presets/` for minion, npc, and boss, indicating different budgets and thresholds can be used for different creature roles.
Typical use cases
Generate a 3D creature from a brief
Use the repository as a text-to-asset pipeline when you want a creature described in prose to become a game-ready GLB without assembling meshes by hand.
Build and validate a controlled workflow
Use the manual, cards, and calibration assets when you need a reproducible process that can be checked for silhouette readability and blocked failures before shipping.
Adapt a documented example to new creature roles
Use the example wolf and role presets as references when adapting the system for lighter quadrupeds, NPC-style creatures, minions, or bosses with different QC thresholds.
Run the compiler directly
Use the engine CLI when you only need the compile step: convert a JSON spec into an output GLB from the command line without adopting the full agent-driven session.
Inspect and package outputs
Use the harness tooling when you want to measure masks, thumbnails, and claims, or generate the offline showroom and delivery pack around the compiled asset.
Pros and Cons
Pros
- Produces a concrete 3D asset from a single JSON spec, with the intended output path and file type clearly documented.
- Combines generation with quality control: calibration files, silhouette gates, and blocked cases are built into the repository.
- Keeps the core engine interface simple, with one command to compile a spec to GLB.
- Documents editable spec controls and role presets, which makes the workflow more inspectable than a black-box prompt-only system.
- Includes a manual and staged cards that explain the generation process step by step.
Cons
- The documentation says the tools for rendering and measurement depend on Playwright and headless Chromium, even though the engine itself has zero runtime dependencies.
- Some capabilities are still described at a high level in the available text, so details about integrations, supported export targets beyond GLB, and broader runtime requirements are not fully specified.
- The README warns that isolated-part blind reads can misfire for certain head shapes, and that older specs may be blocked after breaking-checker changes such as `part_attachment` and `mirror_distortion` in version 1.2.0.
FAQ
What does anyCreature produce?
The repository describes a text-to-creature workflow that starts from a one-sentence order, may ask up to two questions, and then produces a creature from a single JSON spec. The README says the output is a skinned, animated, vertex-coloured, AO-baked GLB plus an offline showroom viewer.
How do you start using it?
The README says to run `bash setup.sh` for dependencies and red/green ruler calibration, then `node engine/cli.js example/wolf.json out/wolf.glb` as the basic compile command. It also says `setup.sh` must print `calibrate OK` before the output can be trusted.
Does it need extra dependencies?
The README states that the engine has zero runtime dependencies and that `node engine/cli.js spec.json out.glb` is the whole interface. The broader harness uses Playwright for rendering and measurement tools, so the tooling side has extra dependencies.
What files document the workflow?
The source says the worked example, `example/wolf.json`, is 2,211 vertices and 31 joints and ships with two animations, not three. It also notes that `cards/` contains the five stage cards the executing session reads in order.
Is there pricing for anyCreature?
The repository includes a LICENSE file and SECURITY.md, but the provided page text does not give a service-style pricing plan for the project itself. As an open-source GitHub repository, there is no product subscription pricing shown on the repo page.
Quick Facts
- Category
- Developer Tool
- Repository host
- GitHub
- Primary output
- Skinned GLB creature file
- Workflow
- Text order → interview → silhouette passes → compile → ship
- Runtime note
- Engine has zero runtime dependencies
- Pricing
- Open-source repository; no product pricing shown
Alternative a anyCreature
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
MeetStream
MeetStream is a meeting bot API for Zoom, Google Meet, Microsoft Teams, and Webex. It helps developers record, stream, and analyze meetings programmatically, with usage-based pricing and a $5 free credit to start.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
hob
hob is an independent workspace for coding agents that keeps agent sessions, terminals, history, and follow-up work organized around the tools and providers you already use. It is aimed at developers who want local control over routing, history, and workspace structure rather than a bundled model stack.
Manta AI
Manta AI is an autonomous web app testing tool for teams that want to map application behavior, catch regressions, and generate tests without writing scripts or maintaining selectors. It works from a URL and supports plain-English test flows, run results with screenshots, and scheduled or deployment-triggered checks.
Redline
Redline is a budgeting tool for Claude Code that paces sessions to stay within time, token, cost, or plan-percentage limits. It uses Claude Code’s native hooks and statusline to help sessions finish with a usable result instead of stopping abruptly.