RealSweatSystem
Sweat, flowing droplets and optional effects for VRChat avatars, with Modular Avatar-based setup. Available on BOOTH and Gumroad with English, Japanese and Korean guides.
Explore productPROJECTS / LAB
Tools, experiments,
and what comes next.
Explore our work across avatar tools, streaming, spatial audio and creative workflows, with the current stage of each project clearly marked.
RealSweatSystem is available to buy. The other entries share our development and research work, with their current stage marked below.
Available now
A released product with store listings and setup documentation.
In development
Tools we are building and validating. Features and release plans may change.
RealWetSystem
A Unity tool in development for the color, sheen and translucency of wet avatar clothing.
Explore productRealRainSystem
Avatar-centered rain effects and their connection to clothing wetness, currently in development.
Explore productCHZZK OSC
A streaming integration tool connecting CHZZK chat, donation and channel-point events to OSC and OBS workflows.
Spatialoom Studio
A local ASMR ambience tool that combines place, weather and time with spatial audio controls. Photo-guided vision API integration is a planned extension, not a released feature.
Cooki Horror Director
A tool for directing VRChat horror-story sessions through avatar menus and OSC, coordinating props, ambience, story stages and sound effects. Currently in release-candidate testing.
Vein Texture Baker
A Unity tool in development for generating vein textures for avatar materials.
DissolveWardrobe
A development preview for avatar first-spawn and outfit-change dissolve transitions, using NDMF-generated FX, menus and material copies.
Prototypes and research
Early implementations for exploring a workflow, not finished products.
PuniBust
An alpha Unity/NDMF tool for non-destructive bust deformation and contact response. It is still being validated and is not a released product.
StreamCut Studio
A local MVP exploring an AI-assisted video-editing workflow for stream recordings.
UpperCapture
An iPhone face, arm and finger tracking prototype with a Windows receiver and optional VMC output for Warudo. Device and avatar validation is still pending.
CookiCam Studio
An OSC camera-direction prototype for VRChat, with framing presets, motion layers and sequences. It depends on a VRCLens-based rig.
RealShaderSystem Patcher
A Unity MVP for material-aware lighting and color refinement on supported Poiyomi and lilToon shaders, using reversible derived-shader patches and before/after previews.
Avatar Outfit Maker
An early research scaffold for natural-language garment design, avatar analysis and fitting validation. End-to-end garment generation is not implemented.
Behind the scenes
Tools for our own production, maintenance and publishing work. They are not offered as public products.
Cooki Wardrobe Studio
An internal garment-photography workflow using real Unity meshes and materials, not AI-generated outfit images.
Facial Tracking Recovery
Internal monitoring, diagnostics and bounded recovery for SRanipal and VRCFaceTracking. Live end-to-end automatic recovery validation is ongoing.
Cooki VOD Title Helper
An internal YouTube Studio helper for preparing titles and descriptions for archived streams.
CookiTools Unity Utilities
Internal Unity utilities for texture optimization, atlas creation, mesh splitting and merging, and batch material editing.
From a scene image
to an editable soundscape.
We have committed to using the Claude API’s vision capabilities for the next stage of our weather and ASMR systems: reading scene clues and turning them into settings people can review and edit.
Choose a reference
A user-selected place photo or game capture provides the visual reference for the scene.
Infer scene clues
Claude will infer the type of place, indoor or outdoor context, visible surface types and weather cues, returning structured settings for review.
Review and build the ambience
People will be able to correct the proposed settings before a local weather and ASMR engine uses them to configure the scene.
A photo provides clues, not measured room dimensions, acoustic material properties or an exact reconstruction of sound. Proposed settings will need human review and listening tests; this is a development plan, not a feature in our released products.