About WPL
The AI governance layer for fitness and wellness.
What is WPL?
WPL (Wellness Plan Language) is the governance layer that ensures AI-generated wellness plans are safe, correct, and free from hallucination. It defines a strict structure for workouts, cardio, nutrition, meditation, recovery, and habits that LLMs must follow.
AI generates plans in a concise WPL-AI DSL, which is then compiled and validated into WPL JSON for storage, rendering, and execution. This pipeline ensures every AI-generated plan is structurally sound before it reaches a client.
WPL brings governance to AI-generated fitness content the way type systems bring safety to code.
Design principles
JSON format parseable by any language, renderable on any platform.
New activity types can be added without breaking existing plans.
Plans adapt based on client input: age, injuries, goals, equipment.
Checkpoints, milestones, and tracking are part of the plan definition.
Plans can include sub-plans and templates. Build once, reuse everywhere.
Built by Gymbile
WPL is created and maintained by Gymbile, a company building the AI governance layer for fitness and wellness.
Gymbile's mission is to make AI-generated wellness plans trustworthy at scale, while keeping them accessible to trainers, coaches, and wellness professionals.
Open standard
WPL is designed as an open standard. Built by Gymbile, open for anyone to use. The full toolchain — specification, JSON Schema, conformance suite, validator, and compiler — is published on GitHub and npm.
- Spec & Schema — github.com/gymbile/wpl · CC-BY-4.0 (spec) + Apache-2.0 (schema)
- Validator — @gymbile/wpl-validator · Apache-2.0
- Compiler — @gymbile/wpl-ai · Apache-2.0
See the ecosystem overview for how the pieces fit together. Issues and contributions welcome.
Get in touch
Questions, feedback, or partnership inquiries? Reach out at hello@gymbile.com.