# Fly trading presentation The articulated fly operates three keys: right foreleg BUY, left foreleg SELL, and both forelegs on the central amber HOLD key. Every recorded native decision serial triggers a 0.7-second press and release, including repeated HOLD. Selected keys stay highlighted afterward. Presentation reads decisions only; it never chooses an action or affects the exchange. Generated walnut workstation, green visor, actual observed AAPL candles, portfolio and native spike monitors. Default Pilot GoPro shows the hands below the wings; Overview still orbits. Direct 1x/4x/12x controls correspond to 10/2.5/0.833 seconds per historical day. Default is 4x. Account/History/Brain tabs consolidate detail; the large latest-action readout shows the committed decision and date, separately from the sampled next signal. Account capital is now $1,000. Show owned share count, its market value, available cash, explicit fill notional, and no-fill reasons. HOLD distinguishes staying in cash from retaining shares. The native texture uses blue ongoing trace and white current spike flags, with dim neutral anatomy. Fixed 0–120 Hz bars show the locked decision sample separately. Periods advance automatically with holdings preserved; final dataset stops. Training has its own clearly labeled trial account and reward, separate from evaluation cash.

# Broker observer camera Attach to the scene's existing player prototype CameraRig. Only its local owner activates camera input. Reuses the published orbit input binding already forked into `roombaCamera`. Right-drag orbit, smoothly normalized gentle wheel zoom, and an under-wing GoPro mount provided by the visual module. UI controls `broker.brokerState.view` and `cameraReset`.

# Historical replay dashboard Displays actual replay dates, explicit seconds per trading day, dollar and percentage P&L, costs, dividends, drawdown, executed orders and comparable passive/random accounts. Native neural activity is a texture; status distinguishes sampling from a completed sample. Data source and train/test split are labelled. No predictive skill is claimed. Compact screens retain tabs, camera controls and a Show UI button.

# Native historical trading experiment Trains the engineered added connections inside the native fly simulation with paired reward trials on 2015 data. Sixteen pairs of 24-day rollouts (768 actual native decisions) compare opposite perturbations of the same weights on identical historical dates and starting cash. Reward is 10000 × log(final equity / initial equity), including commissions, spread, slippage and dividends. No action or inactivity bonus, supervised action labels, forced order, or test-date feedback enters training. The two returns move the 96 shared synapse parameters toward the better perturbation, with bounded updates. This is a small derivative-free reward-search experiment, not proof of predictive skill or a biologically validated learning rule. After an actual decision-neuron silencing check, weights freeze for the entire 2016–17 replay. Native motor spike counts alone select BUY/HOLD/SELL; the exchange may reject an infeasible order but cannot substitute a different action. Starting cash is $1,000. Automatic two-second transitions carry cash, shares, costs, dividends, drawdown and comparison accounts across all five contiguous periods. The final dataset date stops playback and preserves the account; Restart replay explicitly starts over and is marked as repeated evidence. A trading day takes 10/2.5/0.833 seconds at 1x/4x/12x. The decision window locks the action and its sampled rates; native activity continues between decisions without changing that locked action. Pause stops both. Random comparison accounts and a one-time 25% passive allocation use the same exchange assumptions.

# Shared broker presentation state Per-client constants, command queue and runtime references for the Small Brain. Big Trades. experiment. The core drains the queue; the UI and camera keep their own presentation controls here. This module never chooses trades, stimulates neurons or fabricates portfolio values.

# Mesh Mesh asset. `data.zmsh` is the engine-native geometry payload (`renderer.mesh.encode`); consumers reference the asset's guid.

# Texture Managed texture asset. `data.ztex` is the encoded payload the renderer uploads; compression settings live in the `.metadata` `settings` block (edit to re-encode).

# Texture Managed texture asset. `data.ztex` is the encoded payload the renderer uploads; compression settings live in the `.metadata` `settings` block (edit to re-encode).

# Texture Managed texture asset. `data.ztex` is the encoded payload the renderer uploads; compression settings live in the `.metadata` `settings` block (edit to re-encode).