spatial-awareness

SkillAI & models

Introduces a comprehensive agent-based framework for guided software development (GSD)

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the spatial-awareness skill

What this skill tells your AI

The instructions your AI receives, as published by tibsfox/gsd-skill-creator in src/spatial-awareness/SKILL.md and read by ahel’s review.

Passive Sensing, Threat Assessment, and Coordinated Signaling for Agent Systems.

Paula Chipset Release 2 — The Chorus Protocol.

Description

Spatial awareness gives multi-agent systems environmental perception without active polling. The system reads ambient computational signals (context fill, token budget, error rates) and physical sensor data (USB audio, accelerometer, RF spectrum), detects threats through probabilistic scoring with multi-source correlation, and coordinates graduated responses using the 5-phase Frog Protocol.

The output synthesis layer maps sensor data to audio oscillators, DMX-512 lighting, WS2812B LED strips, and ILDA laser projectors — making agent state audible and visible.

Modules

IDModuleFilePurpose
M1Passive Sensorpassive-sensor.tsReads ambient signals, builds spatial model, detects anomalies
M2Threat Detectionthreat-engine.tsProbabilistic scoring, multi-source correlation, sliding window
M3Geometry Mappergeometry-mapper.tsResourceDimension model, constraint boundaries, fill % tracking
M4Frog Protocolfrog-protocol.ts5-phase state machine: BASELINE→SILENCE→ASSESS→PROBE→CLASSIFY→RESUME
M5Comm Buscomm-bus.ts3-tier messaging: COVERT, DIRECTED, BROADCAST
M6Chorus Protocolchorus-proto.tsDistributed pause/resume, scout-first re-engagement, state snapshots

Supporting Layers

LayerFilePurpose
Typestypes.tsZod schemas for all data structures
Sensorssensor-interface.tsUnified SensorStream/OutputDevice contracts
USBusb-device.tsUSB device abstraction (audio, serial, bulk, HID)
Outputoutput-synthesis.tsAudio, DMX, LED, laser output with phase color mapping
Integrationintegration.tsSpatialAwarenessSystem wires full pipeline
Test Envtest-env.ts5-agent SimulatedEnvironment with anomaly injection

Safety-Critical Requirements

IDRequirementStatus
SC-HUMBLOCK threats require human/CAPCOM approval before irreversible actionPASS
SC-RESNo agent resumes work while BLOCK-level threat is active and unapprovedPASS
SC-CORSILENCE entry requires correlated signals from ≥2 sourcesPASS
SC-LASLaser output REFUSED without confirmed interlock (ENGAGED + confirmed)PASS

Core Functionality

IDRequirementStatus
CF-06SILENCE within 500ms of anomaly detectionPASS
CF-07ASSESS produces threat characterization within 5sPASS
CF-08PROBE dispatches scout; scout acts before othersPASS
CF-09CLASSIFY correctly labels THREAT/NEUTRAL/OPPORTUNITYPASS
CF-10RESUME re-engages agents in priority order (scout-first)PASS
CF-11Broadcast delivery to all agents within 1sPASS
CF-12Zero missed broadcastsPASS
CF-13Covert messages invisible to non-recipientsPASS
CF-14Zero work loss on pause/resumePASS
CF-19Audio synthesis from mapped sensorsPASS
CF-20DMX-512 frame output at 44HzPASS
CF-21WS2812B LED strip control with phase colorsPASS
CF-22ILDA laser output with safety interlockPASS

Test Coverage

  • 285 tests across 12 test files
  • 15-scenario threat simulation suite (≥85% TP, ≤10% FP)
  • 50-signal broadcast reliability (zero missed)
  • 10-cycle pause/resume state preservation
  • Full pipeline integration (signal → threat → protocol → output)

Activation

This skill activates when working with:

  • Multi-agent coordination and sensing
  • Threat detection and response protocols
  • Audio/visual output synthesis for agent state
  • Distributed pause/resume with state preservation

Signals

GitHub stars
69
Forks
9
Last commit
Jul 2026

ahel review

  • K1binfo
    installs-packages (in __tests__/types.test.ts)

Automated review, not a security audit. Ruleset v1+k2.

Advanced
Catalog kind
skill
Gateway key
spatial-awareness
Source
github.com/tibsfox/gsd-skill-creator