Agent Sprite Forge

Languages: English | 繁體中文 | 简体中文 | 日本語 | 한국어

0.4 launch teaser: every frame is real ASF output.

One still → a whole game-ready sprite set

Approve one generated still. Get every action of the character, keyed, looped, timed and exported for your engine.

Agent skills for Claude Code (plugin), Codex and Grok. The agent generates a master still with the best image route you have, makes one image-to-video clip per action from it, and deterministic Python tools gate, key, register, loop, retime, finish and package every action for Godot, Aseprite or a web game.

   ![Aria, the approved master still: a young swordswoman with short dark hair, a teal cloak and a long red scarf, side view facing right](./src/v040/aria-master.png) 
    

  1. One approved master still
    

  Generated by the local Codex CLI (image_gen), take 1 of 3.




   ![Aria's six actions at real timing: idle, walk and run loops; attack, jump and hurt one-shots](./src/v040/aria-set.gif) 
    

  2. Six actions, one image-to-video clip each
    

  Grok image-to-video (local CLI), soft-matte keyed, registered, loops picked by measurement, one-shots retimed to game speed, HD finish with the colour lock. Real timing.

3. In a scene: the packaged run, attack and idle over a generated background. The run speed is measured from the contact foot, so the feet do not slide. Every frame comes from the 2026-10-06 live run; nothing is mocked up.

What's new · Quickstart · Routes and providers · Pipeline · Maps · Showcase · Tools

What's new in 0.4 — the big upgrade

Before 0.4, character animation meant cutting frames out of an image-generated sheet, or keying a single clip by hand. 0.4 turns one approved still into the whole set: one image-to-video clip per action, and a measured pipeline from every clip to the engine. Before and after, measured on real outputs:

   ![Old pixel fox run cut from one image-gen sheet next to the new 0.4 pixel fox run, 4x nearest, 80 ms per frame](./src/v040/fox-old-vs-new-4x.gif) 
    

  Pixel finish: the 48x64 fox run
    

  Old: one image-gen sheet cut 1/8 (594–709 colours per frame, soft alpha). New: master still → Grok clip → pixel finish (16–19 colours, binary alpha, selective outline). 4x nearest, both at 80 ms.




   ![2x crop of the same Grok frame: a magenta outline with the old keyer, a clean edge with the 0.4 soft matte](./src/validation/grok-old-vs-new-fringe-2x.png) 
    

  Soft-matte keyer: no purple fringe
    

  Same Grok clip, same frame, 2x crop. Fringe 0.704 → 0, leaked key pixels 7,780 → 0, keying 271 s → 44 s for 145 frames.








   ![Aria's attack at the generator's slow-motion speed next to the auto-retimed attack at game speed](./src/v040/attack-before-after.gif) 
    

  One-shots retimed to game speed
    

  Image-to-video plays a quick attack as a slow-motion clip. retime --auto-oneshot cuts the frozen holds and pins the hit: 4.1 s → 0.7 s, hit at 292 ms.




   ![Aria's run loop without and with the HD colour lock; without it the boots shimmer olive and maroon](./src/v040/colourlock-before-after.gif) 
    

  HD colour lock
    

  Video models drift colours on fast limbs. The lock holds every design colour on the master's (OKLab chroma, lightness kept): hue flips halve and most of the boot shimmer goes away.

What changed:

• Image generation first. Every image and clip goes through one command, route_media.py: your API key if you configured one (OpenAI, Google Gemini, xAI, BytePlus, fal.ai), else your signed-in Codex or Grok CLI, else code art as the last resort. See Routes and providers.
• One master still per character. master_still.py writes the prompt, generates takes, fixes a near-miss by edit and approves one still (master.json).
• A whole set from that still. sprite_set.py makes one image-to-video clip per action, gates every take (area, feet, identity, zoom, turning around, edges, background, extra objects, motion, end pose, colour bleed, timing), retakes with fix clauses, keys with a soft matte, registers, picks loops, retimes one-shots, finishes and packages.
• HD finish by default, crisp pixel finish on request. One body scale per character from the master. The pixel finish adds a crisp downscale, a spaced palette of at most 32 colours, a selective outline and binary alpha.
• Colour lock on every action of a set, and QC recalibrated on 42 labelled live takes.
• Engine exports. animation.json 3.0 with a PNG atlas, VP9-alpha WebM and packed-alpha MP4, decoded and verified; Godot SpriteFrames and Sprite3D; Aseprite JSON; maps for Tiled, Godot and LDtk.
• Claude Code plugin, plus Codex and Grok installs. Hardened on Windows: Codex sandbox run folders, locked-file cleanup, cp1252/cp950 consoles.

Everything in the CHANGELOG. The live record, weak points included: docs/validation-2026-10-06.md.

Quickstart

1. Install

Claude Code (plugin)

claude plugin marketplace add 0x0funky/agent-sprite-forge --sparse .claude-plugin skills
claude plugin install agent-sprite-forge@agent-sprite-forge
python -m pip install "numpy>=1.26" "Pillow>=10.1" "scipy>=1.11"

--sparse checks out only the plugin manifest and the skills, not the showcase media.

Codex

git clone https://github.com/0x0funky/agent-sprite-forge.git
cd agent-sprite-forge
python -m pip install -r requirements.txt
python tools/install_skills.py --apply --host codex

Grok: from the same clone, python tools/install_skills.py --apply --host grok (~/.grok/skills). --host agents writes ~/.agents/skills and --dest any skills folder.

--apply backs up an installed copy and writes a sha256 manifest; python tools/install_skills.py --check --host codex reports drift later. Video needs ffmpeg 5.1+ on PATH. Start a new agent session so the skills load.

Optional: configure an API key (Setting API keys). Without one, the agent uses your signed-in Codex or Grok CLI.

2. Make the master still

Use $generate2dsprite to make a master still of Aria, a young swordswoman hero: short dark hair, a teal hooded cloak, a long red scarf and one steel sword. Side view, facing right, HD.

The agent runs master_still.py generate --takes 3 (every take through route_media.py), shows you the takes, fixes a near-miss with master_still.py edit ("THE ONLY CHANGE: ..."), and approves one: the still is keyed, padded to the class framing (a hero stands 788 px tall on 1024x1024) and recorded in master.json.

3. Make the sprite set

Use $video2dsprite to animate the approved Aria master: idle, walk, run, attack, jump and hurt. Package them for Godot and the web.

Under the hood (`` is the installed video2dsprite folder):

python "/scripts/sprite_set.py" plan --master art/aria-master/master.json --output-dir art/aria-set
python "/scripts/sprite_set.py" run --plan art/aria-set/set_plan.json
python "/scripts/sprite_set.py" review --plan art/aria-set/set_plan.json
python "/scripts/sprite_set.py" accept --plan art/aria-set/set_plan.json --action run --take 2
python "/scripts/sprite_set.py" report --plan art/aria-set/set_plan.json

run resumes where it stopped and never generates a clip twice. The agent looks at every review sheet before it accepts an action, because the numbers cannot see a changed weapon or a covered face. Each action ends as a verified package; export_engine.py adds Godot and Aseprite files.

Routes and providers

Every generated image or clip goes through generate2dmedia/scripts/route_media.py. It walks this order and names the route it used:

python "/scripts/route_media.py" resolve --kind image prints the route that would run, and forge_doctor.py lists which keys are configured (yes or no, never the key) and the resolved order. A moderation refusal, or an outcome that may have cost money, is never retried on another route.

Supported providers

Each model's limits are checked against capabilities.json before anything is sent; a request the model cannot take is skipped with the reason or adapted with a note (for example a video length snapped to one the model renders). Every clip is published without audio. Details: route-media.md, api-usage.md, provider-survey.md, cli-routes.md.

Setting API keys

Each provider reads only its own key, from the environment or from one JSON file in your user configuration folder (never inside a project); the environment wins.

{
  "OPENAI_API_KEY": "sk-...",
  "GEMINI_API_KEY": "...",
  "XAI_API_KEY": "xai-...",
  "ARK_API_KEY": "...",
  "FAL_KEY": "key-id:key-secret",
  "providers": {"order": {"image": ["gemini", "openai"], "video": ["byteplus"]}}
}

Every field is optional. providers.order puts your favourite providers first; models picks a model per provider and tier. Keys are read in-process only: never printed, logged, written, passed on a command line or handed to a child process, and a provider never uses another provider's key. On macOS and Linux keep the file private (chmod 600).

Spend records

Every call, paid or local, is a line in /.forge/ledger.jsonl with its estimate, the provider's job id and the artifact's sha256, and each output folder keeps a job.json. There is no cap unless you set one: --budget-usd, --max-calls, FORGE_MAX_PAID_REQUESTS, and FORGE_SESSION_IMAGES / FORGE_SESSION_VIDEOS for the local routes. media_ledger.py summary shows the totals.

Pipeline: master → set → finish → export

master still        one clip per action      cut                     finish                 package and export
master_still.py --> sprite_set.py run    --> gait_loop.py select --> finish_frames.py   --> engine_export.py package + verify
generate, edit,     route_media.py video,    (loops), retime.py      hd (default) or        export_engine.py (Godot, Aseprite)
approve             gates, retakes, soft     --auto-oneshot          pixel, colour lock
                    matte, registration      (one-shots)             to the master

Engine exports

The pixel fox as shipped: 8 frames, 48x64 cells, 80 ms each, with Godot, Aseprite and web packages (sheet shown at 4x).

Editor imports are not verified yet. The exporters prove parse-level round trips, and Tiled exports are re-rendered at 0 px difference; nobody has opened the 0.4.0 Godot, LDtk or Aseprite exports in the editors.

Maps

generate2dmap makes maps instead of isolated sprites: painted layered maps (a ground-only base, a dressed reference, a prop pack, then anchored transparent props), tiles and autotiles, side-scroll stages and HD-2D battle plates. Collision, reachability and exits come from data (map_bundle.v2, map_nav.py), with exports to Tiled, Godot 4 and LDtk and a single-file walkable HTML preview.

   ![Ground-only cyberpunk canal base map](./src/cyber-canal-base.png) 
    

  Ground-only base




   ![Dressed cyberpunk canal reference map](./src/cyber-canal-dressed-reference.png) 
    

  Dressed reference




   ![Generated 3x3 cyberpunk canal prop pack](./src/cyber-canal-prop-pack.png) 
    

  3x3 prop pack

The layered map: anchored props sorted on their ground lines

Showcase: games made with Agent Sprite Forge

Original games and characters, assembled with Codex and the skills in earlier releases. They show the full loop: generated assets, structured scene data and playable prototype wiring.

   ![Summon Survivors Unity WebGL gameplay](./src/summon-survivors-game-preview1.png) 
    

  Summon Survivors — Unity WebGL
    

  Generated map art, hero sheets, summons, evolutions, enemies, bosses, pickups, HUD, FX, level-up choices and a WebGL build.
    

  [Play build](https://summon-survivors.vercel.app/) · [Build conversation](https://drive.google.com/file/d/1TL7qRX95przTToZILVQ1EFwEXm3flB6t/view?usp=sharing)




   ![Forest Pass Defense Godot tower-defense map](./src/kingdomrush-forest-pass.png) 
    

  Forest Pass Defense — Godot tower defense
    

  A Godot 4 prototype with map, separated props, tower slots, towers, enemy sheets, boss and flying enemies, waves, HUD, build/upgrade/sell flow, projectiles and targeting rules.








   ![Generate2DMap Godot editor scene](./src/godot-editor.png) 
    

  Editable RPG map — Godot TileMap
    

  An image-generated tileset and prop sheet wired into editable TileMapLayer, Sprite2D props, encounter grass Area2D, StaticBody2D collision, exits, metadata and a debug player and camera.




   ![Neon Breach cyberpunk side-scroller](./src/neon-breach.png) 
    

  Neon Breach — cyberpunk side-scroller
    

  A playable side-scroller prototype built around generated character, attack, map and gameplay assets.








   ![Sengoku Era JavaScript RPG starter selection](./src/pokemonlike2.png) 
    

  Sengoku Era — JavaScript monster RPG
    

  A browser RPG prototype with generated characters, starter selection, map flow and battle UI.
    

  [Play build](https://sengoku-era.vercel.app/)




   ![Sengoku Era JavaScript RPG battle scene](./src/pokemonlike.png) 
    

  Starter selection and battle loop
    

  A compact JavaScript game built from sprite, monster, battle and map assets made with the skills.

More Godot tower-defense output

   ![Forest Pass Defense enemy roster](./src/kingdomrush-enemy-roster.png) 
    

  Enemy roster, including flyer and boss units




   ![Forest Pass Defense tower icons](./src/kingdomrush-tower-icons.png) 
    

  Tower lineup




   ![Forest Pass Defense HUD icons](./src/kingdomrush-hud-icons.png) 
    

  HUD and gameplay icons

• scenes/ForestPass.tscn with base map, separated props, enemy paths, tower slots and HUD nodes.
• Six tower families with generated tower art and upgrade stages.
• Animated enemy sheets for ground units, flying units and boss encounters.
• Wave, difficulty, tower catalog, collision, route and tower-slot metadata.
• Runtime build, upgrade, sell, projectile and targeting behaviour connected in Godot.

More Unity survivors-like output

   ![Summon Survivors Unity WebGL gameplay with summons, enemies, pickups, HUD and objective](./src/summon-survivors-game-preview1.png) 
    

  Summons, enemies, pickups, HUD and objective flow




   ![Summon Survivors Unity WebGL level-up menu](./src/summon-survivors-game-preview2-levelup.png) 
    

  Level-up choices: summon unlocks, training, stats and recovery

• Assets/Survivors/Scenes/SummonSurvivors.unity as the playable scene.
• SurvivorContentDatabase.asset connecting generated hero, summon, enemy, pickup, HUD and FX sprites.
• Starter summon selection, survival objective, XP and coin pickups, level-up choices, summon training and evolution flow.
• Enemy spawning pressure, boss timing, projectile attacks, area damage, health bars and score tracking.
• WebGL build output under Builds/WebGL with Vercel deployment config.

Godot editable TileMap: the meadow layers

   ![Godot meadow layered RPG map preview](./src/godot-meadow-layered-preview.png) 
    

  Layered map preview




   ![Godot meadow debug preview with collision and zones](./src/godot-meadow-debug-preview.png) 
    

  Collision and zone debug overlay








   ![Image-generated Godot meadow tileset atlas](./src/godot-meadow-tileset.png) 
    

  Image-generated tileset atlas




   ![Image-generated 3x3 meadow prop pack](./src/godot-meadow-prop-pack.png) 
    

  3x3 generated prop pack

FX and sheet sprites

Sheets stay the route for FX, icons and props: casts, projectiles and impacts come from $generate2dsprite sheets.

   ![Fire mage cast animation](./src/cast.gif) 
    

  Spell cast




   ![Fire mage projectile animation](./src/projectile.gif) 
    

  Projectile

Earlier character output (0.3): four-direction walk sheets and the Ryo video case study

Characters were cut from image-generated sheets before 0.4; 0.4 animates them from one master still instead.

Down

Left

Right

Up

The 0.3 video pipeline on Ryo (base still → 6 s image-to-video → chroma key → 16-frame strip); the 0.4 keyer's numbers on this clip are in the CHANGELOG.

Base still	Motion (image-to-video)	Sprite result (16 frames)
Download MP4

Short skill intro (MP4): intro.mp4

Code art: the last-resort fallback

When no image route exists, or when you ask for code-drawn art, codeart2d draws it from a spec with no image model and no quota, and says so. These are rendered from skills/codeart2d/examples.

   ![Code-drawn pixel hero walk cycle](./src/codeart/hero-walk.gif) 
    

  rig_animate.py: FK/IK walk with planted feet




   ![Code-drawn slash, impact ring, dust and projectile FX](./src/codeart/fx-set.gif) 
    

  fx_build.py: slash, impact, dust and projectile with hit events




   ![Code-made top-down meadow map with roads, a pond, houses and trees](./src/codeart/meadow-layout.png) 
    

  autotile_build.py + layout_build.py: seam-proven tiles, every exit reachable

Included skills

Keep the five skill folders together: they call each other's scripts by relative path.

In Codex, codeart2d and generate2dmedia are explicit-only: the sprite, video and map skills route to them.

Tools

Run from your project root as python "/scripts/.py" .... Every tool has --help, writes a new output folder (it never replaces one), prints one JSON line and exits 1 on failure. The SKILL.md files hold the routing tables.

Requirements

Suggested prompts

A character, start to finish

Use $generate2dsprite to make a master still of a fox ranger for a 2D platformer: orange fur, teal tunic, brown boots, side view facing right. Show me three takes.

Use $video2dsprite to turn the approved fox master into idle, run, attack and hurt, pixel finish in 48x64 cells at 80 ms, and export it for Godot and Aseprite.

Sprites and FX

Use $generate2dsprite to create a wizard spell bundle with cast, projectile and impact sprites.

Use $video2dsprite with my existing side-view hero PNG as the master. Make a run loop and an attack, and report the routes and QC numbers.

Maps

Use $generate2dmap to create a small top-down village with a pond, roads to three exits, collision and a walkable HTML preview, then export it to Tiled.

Use $generate2dmap to create an HD-2D battle plate for a harbour at night with lantern lights and moving water.

Whole-game prompts used for the showcase prototypes

use $generate2dsprite to create a 2D side-scrolling action game. It should include attack mechanics, map elements, and all the essential features. I would like you to design it, and all the necessary assets should be created using this skill. It needs to be an actually playable game, with a cyberpunk story setting.

Use $generate2dsprite to create a 2D monster-collecting RPG. You only need to build one scene for now. It must include a starter monster selection mechanic, a battle screen, and all basic gameplay functions. I would like you to design all the elements and the story, and you can also decide which game engine to use. Use this skill to create any assets you need. The story should be set in the Sengoku period.

Notes and limits

• The scripts are not the creative brain, and numeric QA never approves anatomy or motion by itself: the agent looks at the review sheets, and you approve the master.
• Image-to-video models drift on fast limbs and jumps. Expect retakes on run and jump; the gates and fix clauses exist for that, and the live record shows how many it took.
• What is and is not proven is in docs/known-limitations.md; live results are in docs/validation-2026-10-06.md.
• Engine and editor imports of the 0.4.0 exporters are not verified yet (see Engine exports).

Generated assets and licensing

The MIT license below covers this repository's code and documentation. It does not cover what you generate with it. Images and videos from an image or video model are subject to that provider's terms; code-drawn art is produced from specs you or your agent write. Do not trace or prompt for characters or people you do not own. For commercial projects, use original characters or IP you control, and check each provider's terms.

Repository layout

agent-sprite-forge/
  .claude-plugin/        plugin.json, marketplace.json (Claude Code)
  skills/
    generate2dsprite/    SKILL.md, agents/openai.yaml, references/, scripts/
    video2dsprite/
    generate2dmap/
    generate2dmedia/
    codeart2d/           examples/ with every spec shown above
  shared/                canonical shared modules and JSON schemas (vendored into skills)
  tools/                 install_skills.py, vendor_sync.py, check_links.py
  tests/                 pytest and node suites, fixtures with provenance
  docs/                  validation records, known limitations, audit
  src/                   README media (src/v040: the 0.4 showcase)
  CHANGELOG.md

Star history

License

MIT. See LICENSE.