The Directory 00:00:00
← The directory
TypeScript · React Native · iOS

Cbum Transformation

A local-first iOS training system for programming workouts, logging sets, and measuring readiness against personal health baselines.

Cbum Transformation, a local-first iOS training system for programming workouts, logging sets, and measuring readiness

A gym app with an audience of one.

Cbum Transformation is a personal iOS training system: program the work, log what actually happened, and measure readiness against your own recent baseline. There is no account, analytics pipeline, subscription, or remote workout database.

The app

Today screen
Today
Readiness, streak, next session
Program screen
Program
Twelve weeks at a glance
Training log calendar
Log
Plan beside actual work
Exercise library
Library
1,324 exercises
Workout set logger
Workout
Load, reps, RIR, timer
Completed session summary
Session
Every completed set
Exercise instructions and demo
How-to
Animation and instructions
AI training coach
Coach
Guidance grounded in the log

One source of truth for training

The default plan is Jeff Nippard’s Min-Max Program, parsed from the official spreadsheet into 12 weeks across two blocks, including set-by-set rep ranges, RIR targets, rest intervals, and notes. It can be replaced by a built-in template, imported JSON or CSV, an AI-drafted program, or a program edited in the app.

The calendar keeps planned and completed work separate. A planned rest day and an unplanned lifting session can both be true; the log records both and can sync the result to iCal.

Readiness without false precision

Today’s sleep, HRV, and resting heart rate are compared with the user’s own two-week baseline. Until enough history exists, the app shows no score rather than inventing one. A higher resting heart rate correctly reduces readiness.

Health data and training records remain on the device:

HealthKit ─┐
workouts ──┼─ local SQLite ─ readiness + log + coach context
programs ──┘

The coach receives recent sessions, PRs, and the health trend, and can suggest training adjustments or draft an importable program. Its output is general training guidance, not medical advice.

Stack

The interface uses the project’s Nocturne system: near-black surfaces, electric lime, condensed display type, and monospace numbers for loads, reps, and timers.

Scope

This is a personal app, published as source rather than a consumer product. It is not affiliated with Chris Bumstead, Jeff Nippard, or the exercise-dataset maintainers.