One history. More possibilities.

Switching an interface should not mean abandoning the work you have logged. Training Log Cloud is being designed as a shared backend that authorized apps can read and update through one consistent contract.

Your runs, lifts, swim sessions, and mixed training need to stay understandable across clients. The underlying model keeps future plans separate from actual performance and distinguishes recorded measurements from estimates.

A frontend can emphasize total minutes, lap splits, or strength progress without quietly changing what happened. Derived metrics should explain the records and calculations behind them.

Your data should be able to leave.

Exporting your own training data is a core product requirement. The goal is a useful, structured export you can retain or work with elsewhere, rather than a history trapped in one interface.

Exact export formats, account deletion, retention rules, and supported import paths are still being defined. This page explains the intended controls, not a feature already available for use.

A public frontend and a shared cloud backend are different things. Making HTL's code public does not make athlete data public, or automatically make the backend self-hostable.

Train together.
Choose what travels.

The sharing vision is simple: choose who can see your training, what they can see, and for how long. Your friend should not need the same interface to see the workouts you have chosen to share.

For example, you might share a recent training block with a friend while keeping other sessions private. Viewing and editing need separate permissions. A time range of workouts and the duration of access are also separate choices.

The backend must check current access at the time of every request. Revoking access should stop future reads, although it cannot retrieve a copy someone has already downloaded. Detailed sharing controls remain in development.

An assistant works
within your boundaries.

The planned permission model is per athlete and per capability: Off blocks the operation, Ask first requires approval of a concrete action, and Do it allows that operation within the scope you granted.

Reading a plan, recording a completed workout, and changing your app are different capabilities. Uploading a document should not give an assistant blanket permission to modify your training or share it.

Those boundaries belong on the server. They cannot depend only on a prompt, a button, or a custom frontend behaving correctly. The backend and MCP integration are being developed separately to make that distinction explicit.

Principles first.
Proof before promises.

The platform is in development. Production authentication, sharing behavior, exports, and compatibility need implementation and verification before release. We are not claiming certifications, guaranteed availability, or existing third-party integrations.

See the current development status or learn how the custom frontend concept fits together. For data handling on this marketing site itself, read our website privacy notice.