BraiinsRatchet/START_HERE.md

4.2 KiB

Start Here

This project now has one normal operator entry point:

./scripts/ratchet app

That command builds and opens the native macOS app. The app is the control room. The terminal is only the launcher and fallback diagnostic path.

Your Job

Your job is not to understand every metric.

Your job is:

  1. Open the app with ./scripts/ratchet app.
  2. Stay on Mission Control unless you intentionally need raw diagnostics.
  3. Read Current Decision first.
  4. Read Who Is In Control second.
  5. Use Next Passive Action only when it is enabled.
  6. If you manually place a Braiins canary, record it in Manual Exposure immediately.

Do not start extra terminal watches while the app says a watch, cooldown, or manual exposure owns control.

Who Is In Control?

The app has one ownership model:

  1. The app is ready: you may start the enabled passive action.
  2. A watch run owns control: leave it alone until it finishes.
  3. Cooldown owns control: wait until the shown earliest action time.
  4. Manual exposure owns control: supervise the real-world Braiins/OCEAN position and do not start new experiments.
  5. The app is busy: a monitor-only backend operation is running right now.

This is the anti-babysitting rule: if the app says something else owns control, your workload is zero unless you are supervising a real manual exposure.

What The App Does

The app is monitor-only. It never places, modifies, or cancels Braiins orders.

It can:

  1. Read persisted lifecycle state from data/ratchet.sqlite.
  2. Collect OCEAN and public Braiins market samples.
  3. Run passive watch-only research windows.
  4. Write run reports under reports/.
  5. Track manually executed Braiins exposure that you enter yourself.
  6. Resume from the same SQLite state after a crash or reboot.

Native Mac App

The native SwiftUI app is in:

macos/BraiinsRatchet

Build and open the real app bundle:

./scripts/ratchet app

This creates macos/build/Braiins Ratchet.app. After that, you can open that app bundle directly from Finder or pin it in the Dock.

The app is organized as:

  1. Mission Control: current decision, control ownership, next passive action, progress, evidence, and plain English interpretation.
  2. Research Map: visual autoresearch stage model.
  3. Manual Exposure: record or close manually executed Braiins exposure.
  4. Advanced: raw cockpit, report, and ledger artifacts for diagnostics.
  5. Ratchet Lecture: the general observe, hypothesize, bound, mature, adapt method.

Research Pathway

The app has three time horizons:

  1. Immediate: what can happen now, usually start, wait, refresh, or hold.
  2. Midterm: what probably happens after the current watch, cooldown, or manual exposure matures.
  3. Longterm: what could happen after multiple evidence artifacts point in the same direction.

The pathway is allowed to change after each report. That is the point of ratcheting: the next stage adapts to measured evidence instead of following a rigid plan.

What The Actions Mean

observe means do not bid.

manual_canary means a tiny research experiment is inside the configured loss budget. It is not a profit signal.

manual_bid means the stricter profit-seeking guardrails cleared. The code still does not place the order. You decide manually in Braiins.

Where The Evidence Lives

The master ledger is:

reports/EXPERIMENT_LOG.md

Each completed watch creates one run report:

reports/run-*.md

Use the app's Advanced tab when you need raw artifacts. Mission Control intentionally hides raw logs during normal operation.

Advanced Fallback Commands

Use these only if the native app cannot be opened or you are debugging:

./scripts/ratchet
./scripts/ratchet once
./scripts/ratchet watch 2
./scripts/ratchet supervise
./scripts/ratchet position list
./scripts/ratchet report
./scripts/ratchet experiments

The preferred workflow remains the native app.

The Ratchet Rule

One run is not a verdict. One run is a measurement.

Only change one knob at a time:

  1. Depth target.
  2. Overpay cushion.
  3. Canary spend.
  4. Duration.
  5. Timing window.

Do not increase spend until multiple mature runs point in the same direction.