fable51-worlds README

The project's README from github.com/PhiloLabs/fable51-worlds. As one markdown file: README.md.

fable51-worlds

Worlds as code. A prompt in, a world you can walk out.

Claude Fable 5.1 agent swarms take a brief - a sentence, a photograph, a clip - then research the place, model it, render it, and check it. What ships is a plain Three.js app that opens in a browser.

No game engine. No proprietary 3D tiles. No downloaded meshes. Every building, storefront, sign, tree and traffic light is generated by code that lives in this repo.

Worlds

Every entry is a one-shot result. Where GPT-6 Astra was given the same brief, its world was filmed to the Fable walkthrough's camera route and shot timing, and the two run side by side, unedited - left GPT-6 Astra, right Claude Fable 5.1.

🌉 Union Square, San Francisco

Union Square side by side: GPT-6 Astra on the left, Claude Fable 5.1 on the right

▶ Watch the head-to-head · 59 s · left GPT-6 Astra, right Claude Fable 5.1 · Fable 5.1 build · GPT-6 Astra build · Fable alone: the walkthrough (1920×1080 - aerial, plaza, Nintendo, lower level, Apple)

⛩️ Higashiyama, Kyoto

Kyoto side by side: Codex / GPT-6 Astra on the left, Claude Fable 5.1 on the right

▶ Watch the head-to-head · 53.9 s · left Codex / GPT-6 Astra, right Claude Fable 5.1 · Fable build · Codex source and viewer · Fable alone: the walkthrough (1920×1080 - seven scenes, Gion to Kiyomizu-dera at sunset) · full Codex walkover GIF

🌌 Death Star Trench Run

Trench run side by side: GPT-6 Astra on the left, Claude Fable 5.1 on the right

▶ Watch the head-to-head · 28.2 s · left GPT-6 Astra, right Claude Fable 5.1 · input: a clip from a film · Fable 5.1 build · GPT-6 Astra build · method · Fable alone: the reel (1280×720 - approach, S-foils, interception, dive, trench run, torpedoes, fireball, shockwave) · the brief

More worlds coming.

Any input → a world

Text, video or image. The brief names the subject and the style, and the same pipeline runs behind all three.

Input A brief looks like The world that comes back
📝
Text		 "Hanamikoji, Kyoto - as a hand-painted anime background" A named place in a named style: real geometry on surveyed ground, with the look written to order
🎞️
Video		 a clip from a film The set behind the shot, continued past the edges of frame - step off the camera path and fly the rest of it. See Death Star Trench Run
🖼️
Image		 one photograph, any angle The place in the frame as geometry: turn around, change the hour, light it differently

Style is as open as subject. The same street can come back as a cel-shaded anime plate, a photographic reconstruction or a night scene, because each world's renderer is written for it rather than chosen from a list.

Roadmap

A world is the substrate. What stands on it is the point.
✅ Interactive scenes Real places, walkable in a browser - this repo
🔜 Animation and video Motion as code: shot lists, camera language and continuity held across a sequence minutes long, cut from the world itself rather than generated frame by frame - anime, and documentary camera-matched to the real location
🔜 Agent environments Worlds as training and evaluation environments, where the same code that builds the place also supplies the reward and the verification signal - and, played straight, an RPG
🔜 4D, embodiment, robotics Scenes that change over time, and simulation where the ground truth comes free with the geometry
🔜 Test-time scaling performance Measure how world quality changes with more inference time and compute

How these are made

Every stage is in the repo, and every stage can be re-run:

  1. Reconnaissance - parallel research agents pull map geometry, elevation, transit and street specs, and a storefront census, each fact carrying its source and a confidence level.
  2. Asset generation - scripts emit the kit a world needs: façade modules, street furniture, vehicles, vegetation, fixtures. Some worlds ship no binary assets at all and draw every texture at start-up.
  3. Runtime - a pure Three.js app assembles terrain, streets, façades, props, crowds and traffic from JSON specs.
  4. Verification - Playwright drives the real app, screenshots fixed viewpoints, and diffs them against photographs taken from the same spot. Independent reviewer agents - architect, geographer, technical artist, interaction - file reports that drive the next fix cycle. Builders never grade their own work, and every world ships its own QA report, defects included.

Why code

A world written as code is not the same kind of object as a world generated as pixels or held in a latent space.
Verifiable You can check it by running it. Every elevation in Higashiyama is an independent survey query; a walker drives the real movement code over the whole route; Playwright diffs fixed viewpoints against photographs taken from the same spot. A wrong number is a failing test, not a matter of taste.
Compositional The parts are reusable and legible. A townhouse generator, a roof kit, a street-plot layout engine - each is a module with a contract, and a district is a few hundred lines that calls them.
Editable "Make the shopfront recesses deeper" is a diff, not a re-roll. The world changes exactly where you asked and nowhere else, and the change survives into every later render.

License

Code and generated assets: MIT. Geometry is derived from OpenStreetMap (ODbL), USGS 3DEP (public domain) and the GSI elevation service. Reference photographs are not redistributed here - provenance is recorded per world. Brand names and logos identify the real businesses at their real locations and belong to their owners.

Death Star Trench Run is an unaffiliated homage. Star Wars, X-wing, TIE fighter and the Death Star are trademarks of Lucasfilm Ltd. No film footage, audio, models or textures are used or redistributed here - the reference clip informed the shot list and the look, and every asset in that world is generated procedurally by the code beside it.

fable51-worlds (PhiloLabs/fable51-worlds)

procedural-generation
Searching Database…

523 ·GitHub
Fable51-worlds
What is this?This project turns a single brief - a sentence, a photograph, or a video clip - into a walkable 3D world, generated by Claude-based agent swarms called 'Fable 5.1' that research the real place, model it, render it, and verify it.
Is this a game engine, or something else?It is not a game engine: every world ships as a plain Three.js web app, with no proprietary 3D tiles and no downloaded meshes, since every building, sign, and tree is generated by code in the repo.
How are these worlds actually built and re-run?Every stage can be re-run from the repo: reconnaissance agents pull map and elevation data, asset-generation scripts emit façades and props, a Three.js runtime assembles the scene from JSON specs, and Playwright-driven verification diffs screenshots against real photographs.
What powers the pipeline - Claude, Three.js, something else?The pipeline is built on Claude agent swarms for research and generation, a pure Three.js runtime for rendering, Playwright for automated verification, and real-world geometry pulled from OpenStreetMap, USGS 3DEP, and the GSI elevation service.
Is the whole roadmap already working, or just part of it?Only the interactive-scenes stage is live right now; animation/video, agent training environments, 4D/embodiment/robotics simulation, and test-time scaling benchmarks are all listed as upcoming roadmap items, not shipped features.
Do I need a game engine or special hardware to view a world?No - the generated worlds open directly in a standard browser via WebGL and Three.js, with no proprietary engine, installer, or downloaded assets required.
What's the licensing on the code and the generated worlds?Code and generated assets are MIT licensed, but the underlying geometry is derived from OpenStreetMap (ODbL), USGS 3DEP (public domain), and the GSI elevation service, and reference photographs are not redistributed.
What is Fable 5.1 exactly?It is the name of the Claude-based agent swarm system that researches, models, renders, and verifies each generated world, and it is directly compared head-to-head against GPT-6 Astra on identical briefs.
How does the verification step actually check a world?Playwright drives the running app, screenshots fixed viewpoints, and diffs them against real photographs taken from the same spot; independent reviewer agents (architect, geographer, technical artist, interaction) file QA reports that feed the next fix cycle.
Can I feed it a video clip instead of just text?Yes - inputs can be text, a single photograph, or a video clip, as shown by the Death Star Trench Run example, which extrapolated the full set behind a film shot from one reference clip.
Why build worlds as code instead of generating pixels or video?Code is verifiable (you can run it and check it against reality), compositional (reusable modules like façade and roof generators), and editable (a request like deeper shopfront recesses becomes a precise diff rather than a full re-roll).
What are similar projects?natural-disasters, threejs-architecture-effects, LinearAbilityExtThreeJS, dream-loop, genart-skill, image-to-3d-pipeline, anyCreature, Hermes3D
Which topics is it in?openstreetmap, 3d-graphics, procedural-generation, 3d-reconstruction, world-models, threejs, simulation, webgl, playwright, game-development, 3d, agentic-ai
Creator
PhiloLabs
Size
416.7 MB
Stars
523
Updated
CommitSep 09
Age
1mo
Commits · 90 Days