Skip to content

johannes42x/trainingOS

Repository files navigation

trainingOS

This repository is part of MetaMemoryWorks
a file-based architecture for persistent AI memory.

🌐 http://metamemoryworks.com
📄 Architecture: https://github.com/johannes42x/MetaMemoryArchitecture

A file-based, persistent training decision system for safe, adaptive, reasoning-driven fitness assistance.

Version: 0.9


What is trainingOS?

trainingOS is a method and product for supporting physical training (strength, endurance, recovery) using persistent, file-based logs and reasoning-based decision logic.

Instead of fixed plans or opaque app logic, trainingOS works with:

  • real training logs
  • subjective states (fatigue, pain, sleep quality)
  • safety-first constraints
  • long-term pattern recognition

to provide adaptive, explainable training decisions.

trainingOS is part of the MetaMemoryWorks (MMW) ecosystem, but can also be used as a standalone module.


Core Capabilities

  • Persistent evaluation of training logs (strength, cardio, recovery)
  • Safety-first logic (pain, exhaustion, plausibility checks come first)
  • Distinction between real progress and compensatory overload
  • Detection of implausible or inconsistent data
  • Long-term pattern recognition across sessions
  • Clear, text-based guidance with explicit uncertainty handling

trainingOS prioritizes injury prevention, sustainability, and honesty over short-term performance optimization.


What trainingOS is not

  • Not a click-based fitness app
  • Not a static training plan generator
  • Not a medical diagnostic system
  • Not a wearable replacement
  • Not a motivational hype engine

trainingOS does not give medical advice and does not override pain or safety signals.


Method & Architecture

trainingOS follows a file-based, persistent OS paradigm:

  • user data lives in explicit files (logs, states, presets)
  • decisions reference historical context
  • uncertainty is preserved, not filled with guesses
  • safety constraints cannot be disabled by user prompts

The protected scope includes both:

  • the method (decision logic, safety model, persistence pattern)
  • and products implementing it

Renaming components or changing formats does not remove licensing obligations if the functional method is reproduced.


For licensing inquiries, contact: lizenz@metamemoryworks.de / licensing@metamemoryworks.com

For everything else: kontakt@metamemoryworks.de / contact@metamemoryworks.com

Johannes Glaser
Author & Rights Holder
MetaMemoryWorks

About

A stateful training assistant built on persistent memory — plans, logs, and long-term progression instead of ephemeral chat context.

Topics

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors