What Happened

A 3D desktop — driven by a live spiking simulation FlyWire connectome. It walks across your windows, grooms, sleeps, and decides cursor neurons a .

markers (AGENTS.md, CLAUDE.md, .cursor/rules, .kiro/steering and ~40 more) and turns anything on screen source: an editor or terminal window with the project open, a Finder window, a folder desktop. An open project smells strongest, a closed icon weakest, vibecode it holds — a marked repos is smelled across the whole screen, a single weak folder only from nearby. The steering neurons , population wakes up enough .

The fly's brain window: 23,210 real neuron soma positions from FlyWire v783, with live spikes flashing at real neuron locations. The two glowing yellow markers Fibers — the escape command neurons. Click any region to stimulate it.

23,210 neuron soma positions (of 139,255 in FlyWire v783) render the rotating brain window, colored by super-class (FlyWire's coarse cell-type grouping). A 668-neuron circuit with ~19,000 real synaptic connections (synapse counts, signed by neurotransmitter prediction) runs as a 1 kHz leaky-integrate-and-fire (LIF) simulation: LC4 (104) + LPLC2 (210) looming-detector visual neurons DNp01 / Giant Fiber (GF) (2) — the escape command neuron DNa01 + DNa02 (4) steering neurons · DNp09 (2) forward walking DNg11 (6) grooming · MDN (4) backward walking ("moonwalker") DNp02/DNp04/DNp11 (6) escape-maneuver (wing) neurons their 330 strongest partners, including ascending (proprioceptive) and sensory (wind) neurons Escape is not scripted. Your cursor's approach becomes looming LC4/LPLC2 cells; actually spikes through its real synapses — ~1,200 synapses of feedforward inhibition push back, approaches are tolerated and fast lunges trigger escape in ~4 ms, animal. The body itself is procedural (FlyWire is a brain connectome — no body geometry exists), with a tripod gait, visible wing-beat, altitude-scaled flight, grooming, and sleep postures.

Why It Matters

Requirements: macOS 13+, Xcode Command Line Tools (Swift 5.9+). No permissions or entitlements needed — everything it senses (cursor, window frames, clicks-as-taps, thermal state) is permission-free.

://github.com/DenisSergeevitch/desktop-fly.git cd desktop-fly ./build.sh ./DesktopFly A 🪰 item appears ; . The fly wanders your desktop on a transparent, click-through overlay — it never intercepts keyboard.

item effect Pause / Resume freeze /Hide Brain toggle window Escape Test (loom) inject a looming stimulus, Display across monitors (shown when >1 display) Add / Remove (only fly #1 carries the brain) Scare Flies startle everyone The brain window is interactive: hovering pauses the rotation; clicking a region "optogenetically" stimulates the ~60 nearest circuit neurons for 400 ms. The fly's reaction is whatever the real network does downstream — escapes; click DNg11 and it grooms; 's DNa01/02 .

body behavior driven by escape takeoff DNp01 . rest, walking speed DNp09 rate steering DNa01+DNa02 left−right rate difference grooming DNg11 rate backward nervous darting LC4/LPLC2 population rate wing-beat effort, threat wing-raise DNp02/04/11 rate spontaneous takeoff whole-population arousal closes body→brain: the gait rhythm feeds the circuit's real ascending (proprioceptive) neurons , and fast cursor motion stimulates its sensory (wind) partners.

What Comes Next

Window terrain: window ledges — , , rides a window you drag, and startles when one closes . Window looms: a window appearing looming pathway; the circuit decides whether dialogs. Clicks are substrate taps; clicking startles it through the wind→GF pathway. Typing is vibration (idle-time API — pressed, never which). Circadian rhythm: dawn/dusk activity peaks, midday siesta, night quiescence. Sleep: → it sleeps, breathing slowly, with raised arousal threshold; it grooms after waking. Temperature: flies are ectotherms — a a faster fly. Regenerating / ships with compact derived files. To rebuild FlyWire Codex dumps (~60 MB download):

mkdir -p /tmp/flywire && cd /tmp/flywire B=https://storage.googleapis.com/flywire-data/codex/data/fafb/783 curl -O "$B/classification.csv.gz" -O "$B/coordinates.csv.gz" \ -O "$B/connections.csv.gz" -O "$B/consolidated_cell_types.csv.gz" cd - && python3 etl.py /tmp/flywire Diagnostics ./DesktopFly --simtest # circuit invariants: GF silent at rest, 4 ms loom latency, ... ./DesktopFly --behaviortest # 17 end-to-end checks: stimulate neurons -> body reacts ./DesktopFly --snapshot f.png # offscreen fly render ./DesktopFly --brainshot b.png # offscreen brain render What's modeled vs. measured Honesty section: the connectome gives wiring, not physiology. The LIF dynamics, neurotransmitter signs (ACh+, GABA−, Glu−), the gap-junction →→GF (documented electrical coupling), synaptic delays, and the sensory transduction (cursor → looming value) are standard modeling choices layered . Everything downstream of the sensory neurons — who connects to whom, and how strongly — is FlyWire data.

. / are derived from FlyWire (FAFB v783) -NC 4.0 — see data/DATA_LICENSE.md. , cite:

Explore more: Business Plan Startup Strategy · Micro Saas Bootstrapping