Posecode

Kinematic motion as text. Mermaid gave LLMs a way to draw diagrams.
Posecode gives them a way to show movement: exercises, physiotherapy, posture,
as a tiny human-readable language that renders to an animated 3D figure in the browser.

Live playground · Movement library · Language spec · Examples · MCP server

_{One .posecode document — shoulders: abduct 160, hips: abduct 30, repeat 12 — rendered live in the browser.}

pelvis: hinge, deadlift

knees: flex 95, squat

shoulders: abduct 90, lateral raise

Ask an LLM to explain a physical movement, and it will give you unstructured prose or a static, flat diagram. But large language models already understand the biomechanics of movement (e.g., "elbows flex, shoulders abduct on the descent of a push-up"). They just lack a standardized syntax to express it in a way that a computer can render dynamically.

While neural network-based text-to-motion models exist, they are impractical for consumer web applications:

• Resource Intensive: They require heavy, expensive GPU hosting, making real-time generation and scaling cost-prohibitive.
• No Fine Control: They output black-box 3D coordinate trajectories, making it impossible to adjust anatomical phases, joint limits, or speed programmatically.
• Safety Hazards: There are no safety boundaries, meaning the model can easily render joint extensions that are anatomically impossible or physically dangerous.

Posecode takes the opposite, lightweight approach:

• Text-Driven: The LLM writes a tiny .posecode text document specifying semantic joint angles and phase times—generation costs a fraction of a cent.
• Unbelievably Fast: A client-side parser and WebGL renderer animate the figure at 60 FPS directly in the browser—even on low-end mobile devices.
• Anatomically Safe: Every joint rotation is clamped to clinical range-of-motion limits from standard physiotherapy tables. Hallucinations like knee: flex 200 are safely capped with warnings.

A .posecode file describes human movements as a sequence of timed steps with targeted joint movements and range-of-motion rules:

1. Write .posecode	2. Render 3D Animation
posecode exercise "Body-weight squat"   rig humanoid   pose start = standing   step "Descend" 1.6s ease-in-out:     hips: flex 80     knees: flex 95     ankles: dorsiflex 14     ground-lock: feet     cue "Sit the hips back..."   step "Drive up" 1.2s ease-out:     hips: flex 0     knees: flex 0     ankles: dorsiflex 0     ground-lock: feet     cue "Drive through the heels..."   repeat 8

Choose the integration path that fits your use case:

Instantly preview, edit, and share movements in the browser: posecode.org/play

Teach your AI agent (in Claude Desktop, Cursor, etc.) to read, write, and render Posecode natively using our Model Context Protocol server:

# Add to your MCP client config (e.g. claude_desktop_config.json):
npx posecode-mcp

See the MCP Package README for full configuration options.

Embed an interactive, low-poly 3D player on any page using a single <script> tag:

<script src="https://unpkg.com/posecode-embed/dist/posecode-embed.js"></script>

<posecode-player src="/movements/squat.posecode"></posecode-player>

See the Embed Package README for customizing autoplay, controls, speed, and styling.

Build custom rendering or parsing logic directly in your own applications:

# Parser only (converts text to range-of-motion clamped IR)
npm install posecode-parser

# WebGL 3D Renderer (built on Three.js)
npm install posecode-render

Two safety layers ship with the language:

• ROM clamping: every angle is hard-clamped to healthy range-of-motion tables before rendering; a hallucinated knee: flex 200 renders at its ceiling with a warning, never an impossible joint.
• Fidelity evals: posecode-eval re-runs the real parser → FK → ground-lock pipeline headlessly and scores geometric invariants ("a deadlift pitches the torso ≥ 50° with vertical shins"). Every example must pass every invariant in CI.

The protocol and both libraries are MIT-licensed: the open core. See spec/SPEC.md for the full language and spec/llm-authoring.md for the authoring prompt. For where Posecode spreads fastest and the per-domain go-to-market plan, see docs/market-research.md; for the engine roadmap, ROADMAP.md.

The 3D figure is the Adobe Mixamo anatomical mannequin character, exported from Mixamo and used under the Mixamo license (royalty-free in projects). The renderer also ships a zero-asset procedural figure, used automatically wherever the character can't load — and it accepts any Mixamo-rigged GLB via characterUrl.

Single-person movement across fitness, physio, desk, dance, education & rehab · Mermaid-style DSL · ROM safety clamping (authored and IK-solved angles) · forward kinematics · ground-lock and ROM-constrained reach-to-target IK · hip-hinge · lying/seated poses · scene props (chair/wall/bar) · a single-DOF hand rig · live playground.

Deferred: two-person / partner movements + collision detection, deeper props (load, bands, rings), multi-joint fingers, FBX/GLB export, hosted SaaS editor and the expert-verified motion marketplace.

This project follows a design study, "Kinematic Motion Definition Protocols for Large Language Models", which argues for a semantic DSL over diffusion models, specifies ROM-based safety constraints from clinical normative data, and lays out the open-core commercialization path. The spec cross-references its sections (§4 DSL, §5 biomechanics, §6 client rendering, §7 strategy).

Posecode's range-of-motion values are general literature data, not medical advice. Consult a qualified professional for physiotherapy or exercise prescription.

We'd love to hear your feedback! You can reach us in two ways:

• Email: Send us an email at hello@posecode.org.
• GitHub Issues: If you found a bug or have a feature request, please open a GitHub Issue.