feat(yolo-coral-tpu): migrate completely to native inline OS GUI driver deployments

- Drop all Docker architecture dependencies
- Remove Aegis manual_setup firewall pre-conditions
- Implement native ask_sudo wrapped in osascript/pkexec for UI collection
- Fully script bare-metal Windows UAC powershell payload handling

Author: solderzzc

skills/detection/yolo-detection-2026-coral-tpu/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: yolo-detection-2026-coral-tpu
-description: "Google Coral Edge TPU — real-time object detection via Docker"
+description: "Google Coral Edge TPU — real-time object detection natively via local Python environment"
 version: 1.0.0
 icon: assets/icon.png
 entry: scripts/monitor.js
@@ -10,18 +10,7 @@ runtime: node
 requirements:
   platforms: ["linux", "macos", "windows"]
 
-manual_setup:
-  macos: |
-    # 1. Fix Homebrew permissions and install libusb
-    sudo chown -R $(whoami) /opt/homebrew
-    brew install libusb
 
-    # 2. Add Apple Silicon Edge TPU native driver
-    curl -sSLO https://github.com/feranick/libedgetpu/releases/download/16.0TF2.19.1-1/libedgetpu-16.0-tf2.19.1-1_MacOS_Silicon.zip
-    unzip -q -o libedgetpu-16.0-tf2.19.1-1_MacOS_Silicon.zip
-    sudo mkdir -p /usr/local/lib
-    sudo cp libedgetpu.1.dylib /usr/local/lib/
-    rm -rf libedgetpu*
 
 parameters:
   - name: auto_start
@@ -90,13 +79,13 @@ mutex: detection
 
 # Coral TPU Object Detection
 
-Real-time object detection using Google Coral Edge TPU accelerator. Runs inside Docker for cross-platform support. Detects 80 COCO classes (person, car, dog, cat, etc.) with ~4ms inference on 320×320 input.
+Real-time object detection natively utilizing the Google Coral Edge TPU accelerator on your local hardware. Detects 80 COCO classes (person, car, dog, cat, etc.) with ~4ms inference on 320x320 input.
 
 ## Requirements
 
 - **Google Coral USB Accelerator** (USB 3.0 port recommended)
-- **Docker Desktop 4.35+** (all platforms — Linux, macOS, Windows)
-- USB 3.0 cable (the included cable is recommended)
+- **libusb** framework (installed automatically on Linux/macOS)
+- Python 3 with the native `pycoral` environment
 - Adequate cooling for sustained inference
 
 ## How It Works
@@ -106,61 +95,50 @@ Real-time object detection using Google Coral Edge TPU accelerator. Runs inside
 │ Host (Aegis-AI)                                     │
 │   frame.jpg → /tmp/aegis_detection/                 │
 │   stdin  ──→ ┌──────────────────────────────┐       │
-│              │ Docker Container              │       │
+│              │ Native Python Environment     │       │
 │              │   detect.py                   │       │
 │              │   ├─ loads _edgetpu.tflite     │       │
-│              │   ├─ reads frame from volume   │       │
+│              │   ├─ reads frame from disk     │       │
 │              │   └─ runs inference on TPU    │       │
 │   stdout ←── │   → JSONL detections          │       │
 │              └──────────────────────────────┘       │
-│   USB ──→ /dev/bus/usb (Linux) or USB/IP (Mac/Win) │
+│   USB ──→ Native System USB / edgetpu drivers       │
 └─────────────────────────────────────────────────────┘
 ```
 
-1. Aegis writes camera frame JPEG to shared `/tmp/aegis_detection/` volume
-2. Sends `frame` event via stdin JSONL to Docker container
-3. `detect.py` reads frame, runs inference on Edge TPU
+1. Aegis writes camera frame JPEG to shared `/tmp/aegis_detection/` workspace
+2. Sends `frame` event via stdin JSONL to the local Python instance
+3. `detect.py` invokes PyCoral and executes natively on the mapped USB Edge TPU
 4. Returns `detections` event via stdout JSONL
-5. Same protocol as `yolo-detection-2026` — Aegis sees no difference
 
 ## Platform Setup
 
 ### Linux
 ```bash
-# USB Coral should be auto-detected
-# Docker uses --device /dev/bus/usb for direct access
+# Uses the official apt-get google-coral packages natively
 ./deploy.sh
 ```
 
-### macOS (Docker Desktop 4.35+)
+### macOS 
 ```bash
-# Docker Desktop USB/IP handles USB passthrough
-# ARM64 Docker image builds natively on Apple Silicon
+# Downloads and installs the libedgetpu OS payload framework inline
 ./deploy.sh
 ```
 
 ### Windows
 ```powershell
-# Docker Desktop 4.35+ with USB/IP support
-# Or WSL2 backend with usbipd-win
+# Installs directly to the Microsoft runtime
 .\deploy.bat
 ```
 
-## Model
-
-Ships with pre-compiled `yolo26n_edgetpu.tflite` (YOLO 2026 nano, INT8 quantized, 320×320). To compile custom models:
-
-```bash
-# Requires x86_64 Linux or Docker --platform linux/amd64
-python scripts/compile_model.py --model yolo26s --size 320
-```
+> **Important Deployment Notice**: The updated `deploy.sh` script will natively halt execution and prompt you securely for your OS `sudo` password to securely register the USB drivers (`libedgetpu`) system-wide. If you refuse the prompt, it gracefully outputs the exact terminal instructions for you to configure it manually.
 
 ## Performance
 
 | Input Size | Inference | On-chip | Notes |
 |-----------|-----------|---------|-------|
-| 320×320 | ~4ms | 100% | Fully on TPU, best for real-time |
-| 640×640 | ~20ms | Partial | Some layers on CPU (model segmented) |
+| 320x320 | ~4ms | 100% | Fully on TPU, best for real-time |
+| 640x640 | ~20ms | Partial | Some layers on CPU (model segmented) |
 
 > **Cooling**: The USB Accelerator aluminum case acts as a heatsink. If too hot to touch during continuous inference, it will thermal-throttle. Consider active cooling or `clock_speed: standard`.
 
@@ -184,4 +162,4 @@ Same JSONL as `yolo-detection-2026`:
 ./deploy.sh
 ```
 
-The deployer builds the Docker image locally, probes for TPU devices, and sets the runtime command. No packages pulled from external registries beyond Docker base images and Coral apt repo.
+The deployer builds the local native Python virtual environment inline with global TPU hooks. No Docker containers or abstract container-bindings are used.

skills/detection/yolo-detection-2026-coral-tpu/deploy.bat

@@ -1,67 +1,147 @@
 @echo off
-REM deploy.bat — Docker-based bootstrapper for Coral TPU Detection Skill (Windows)
+REM deploy.bat — Native local bootstrapper for Coral TPU Detection Skill (Windows)
 REM
-REM Builds the Docker image locally and verifies Edge TPU connectivity.
-REM Called by Aegis skill-runtime-manager during installation.
+REM Builds a local Python virtual environment and installs the Google Coral Edge TPU
+REM natively on the host OS. Safely triggers UAC elevation for driver installation.
 REM
-REM Requires: Docker Desktop 4.35+ with USB/IP support
+REM Exit codes:
+REM   0 = success
+REM   1 = fatal error (Python/pip not found or UAC denied)
+REM   2 = partial success (no TPU detected, will use CPU fallback)
 
 setlocal enabledelayedexpansion
 
 set "SKILL_DIR=%~dp0"
-set "IMAGE_NAME=aegis-coral-tpu"
-set "IMAGE_TAG=latest"
 set "LOG_PREFIX=[coral-tpu-deploy]"
 
-REM ─── Step 1: Check Docker ────────────────────────────────────────────────
+REM Ensure we run inside the correct folder
+cd /d "%SKILL_DIR%"
 
-where docker >nul 2>&1
+echo %LOG_PREFIX% Platform: Windows 1>&2
+echo {"event": "progress", "stage": "platform", "message": "Windows native environment detected"}
+
+REM ─── Step 1: Install Native OS TPU Drivers (UAC Promoted) ───────────────
+
+echo %LOG_PREFIX% Downloading Google official x64 Windows installer... 1>&2
+
+set "TMP_DIR=%TEMP%\coral_tpu_install"
+if not exist "%TMP_DIR%" mkdir "%TMP_DIR%"
+cd /d "%TMP_DIR%"
+
+powershell -Command "Invoke-WebRequest -Uri 'https://github.com/google-coral/libedgetpu/releases/download/release-grouper/edgetpu_runtime_20221024.zip' -OutFile 'edgetpu_runtime_20221024.zip'"
 if %errorlevel% neq 0 (
-    echo %LOG_PREFIX% ERROR: Docker not found. Install Docker Desktop 4.35+ 1>&2
-    echo {"event": "error", "stage": "docker", "message": "Docker not found. Install Docker Desktop 4.35+"}
+    echo %LOG_PREFIX% ERROR: Failed to download Edge TPU runtime. 1>&2
     exit /b 1
 )
 
-docker info >nul 2>&1
+powershell -Command "Expand-Archive -Path 'edgetpu_runtime_20221024.zip' -DestinationPath '.' -Force"
+cd edgetpu_runtime
+
+echo %LOG_PREFIX% Prompting for Administrator rights to install drivers... 1>&2
+echo {"event": "progress", "stage": "platform", "message": "UAC required: Install Google Coral Edge TPU Windows Runtime"}
+
+REM Start the official install script elevated. Wait for it to finish.
+powershell -Command "Start-Process -FilePath 'cmd.exe' -ArgumentList '/c', 'install.bat < nul' -Verb RunAs -Wait"
+
 if %errorlevel% neq 0 (
-    echo %LOG_PREFIX% ERROR: Docker daemon not running. Start Docker Desktop. 1>&2
-    echo {"event": "error", "stage": "docker", "message": "Docker daemon not running"}
+    echo.
+    echo [MANUAL SETUP REQUIRED] UAC prompt was skipped or user aborted.
+    echo Please execute the following fragile instructions manually in an Administrator console:
+    echo.
+    echo powershell -Command "Invoke-WebRequest -Uri 'https://github.com/google-coral/libedgetpu/releases/download/release-grouper/edgetpu_runtime_20221024.zip' -OutFile 'edgetpu_runtime_20221024.zip'"
+    echo powershell -Command "Expand-Archive -Path 'edgetpu_runtime_20221024.zip' -DestinationPath '.'"
+    echo cd edgetpu_runtime
+    echo install.bat
+    echo.
+    echo Once completed, re-run this deployment.
+    echo {"event": "error", "stage": "platform", "message": "Manual OS setup required (UAC skipped)"}
+    cd /d "%SKILL_DIR%"
+    rmdir /S /Q "%TMP_DIR%"
     exit /b 1
 )
 
-for /f "tokens=*" %%v in ('docker version --format "{{.Server.Version}}" 2^>nul') do set "DOCKER_VER=%%v"
-echo %LOG_PREFIX% Using Docker (version: %DOCKER_VER%) 1>&2
-echo {"event": "progress", "stage": "docker", "message": "Docker ready (%DOCKER_VER%)"}
+cd /d "%SKILL_DIR%"
+rmdir /S /Q "%TMP_DIR%"
 
-REM ─── Step 2: Build Docker image ──────────────────────────────────────────
+REM ─── Step 2: Ensure Python 3 ──────────────────────────────────────────────
 
-echo %LOG_PREFIX% Building Docker image: %IMAGE_NAME%:%IMAGE_TAG% ... 1>&2
-echo {"event": "progress", "stage": "build", "message": "Building Docker image..."}
+set "PYTHON_CMD=python"
+python --version >nul 2>&1
+if %errorlevel% neq 0 (
+    python3 --version >nul 2>&1
+    if !errorlevel! equ 0 (
+        set "PYTHON_CMD=python3"
+    ) else (
+        echo %LOG_PREFIX% ERROR: Python not found. 1>&2
+        echo {"event": "error", "stage": "python", "message": "Python not found"}
+        exit /b 1
+    )
+)
+
+echo %LOG_PREFIX% Using Python: !PYTHON_CMD! 1>&2
+
+REM ─── Step 3: Create Virtual Environment ───────────────────────────────────
 
-docker build -t %IMAGE_NAME%:%IMAGE_TAG% "%SKILL_DIR%"
+set "VENV_DIR=%SKILL_DIR%venv"
+echo %LOG_PREFIX% Setting up virtual environment in %VENV_DIR%... 1>&2
+echo {"event": "progress", "stage": "build", "message": "Creating Python virtual environment..."}
+
+!PYTHON_CMD! -m venv "%VENV_DIR%"
+
+if not exist "%VENV_DIR%\Scripts\python.exe" (
+    echo %LOG_PREFIX% ERROR: Failed to create virtual environment. 1>&2
+    echo {"event": "error", "stage": "build", "message": "Failed to create venv"}
+    exit /b 1
+)
+
+REM ─── Step 4: Install PyCoral & Dependencies ───────────────────────────────
+
+echo %LOG_PREFIX% Installing Python dependencies (including pycoral)... 1>&2
+echo {"event": "progress", "stage": "build", "message": "Fetching pycoral specific wheels..."}
+
+"%VENV_DIR%\Scripts\python.exe" -m pip install --upgrade pip >nul 2>&1
+
+"%VENV_DIR%\Scripts\python.exe" -m pip install --extra-index-url https://google-coral.github.io/py-repo/ pycoral~=2.0
 if %errorlevel% neq 0 (
-    echo %LOG_PREFIX% ERROR: Docker build failed 1>&2
-    echo {"event": "error", "stage": "build", "message": "Docker image build failed"}
+    echo %LOG_PREFIX% WARNING: pip install pycoral failed. Will attempt to install standard requirements anyway. 1>&2
+)
+
+"%VENV_DIR%\Scripts\python.exe" -m pip install -r "%SKILL_DIR%requirements.txt"
+if %errorlevel% neq 0 (
+    echo %LOG_PREFIX% ERROR: Failed to install Python dependencies. 1>&2
+    echo {"event": "error", "stage": "build", "message": "pip install failed"}
     exit /b 1
 )
 
-echo {"event": "progress", "stage": "build", "message": "Docker image ready"}
+echo %LOG_PREFIX% Dependencies installed successfully. 1>&2
+
+REM ─── Step 5: Probe for Edge TPU devices ────────────────────────────────────
+
+echo %LOG_PREFIX% Probing for Edge TPU devices natively... 1>&2
+echo {"event": "progress", "stage": "probe", "message": "Checking for physical Edge TPU..."}
 
-REM ─── Step 3: Probe for Edge TPU ──────────────────────────────────────────
+set "TPU_FOUND=false"
+set "TPU_COUNT=?"
 
-echo %LOG_PREFIX% Probing for Edge TPU devices... 1>&2
-echo {"event": "progress", "stage": "probe", "message": "Checking for Edge TPU devices..."}
+for /f "delims=" %%I in ('"%VENV_DIR%\Scripts\python.exe" "%SKILL_DIR%scripts\tpu_probe.py" 2^>nul') do (
+    set "PROBE_OUTPUT=%%I"
+)
 
-docker run --rm --privileged %IMAGE_NAME%:%IMAGE_TAG% python3 scripts/tpu_probe.py >nul 2>&1
+echo !PROBE_OUTPUT! | findstr /C:"\"available\": true" >nul
 if %errorlevel% equ 0 (
-    echo %LOG_PREFIX% Edge TPU detected 1>&2
-    echo {"event": "progress", "stage": "probe", "message": "Edge TPU detected"}
+    set "TPU_FOUND=true"
+    REM Approximate counts natively
+    echo {"event": "progress", "stage": "probe", "message": "Edge TPU device natively registered"}
 ) else (
-    echo %LOG_PREFIX% WARNING: No Edge TPU detected - CPU fallback 1>&2
-    echo {"event": "progress", "stage": "probe", "message": "No Edge TPU detected - CPU fallback"}
+    echo {"event": "progress", "stage": "probe", "message": "No Edge TPU detected -- CPU fallback available"}
 )
 
-echo {"event": "complete", "status": "success", "message": "Coral TPU skill installed"}
+REM ─── Step 6: Complete ──────────────────────────────────────────────────────
 
-echo %LOG_PREFIX% Done! 1>&2
-exit /b 0
+if "!TPU_FOUND!"=="true" (
+    echo {"event": "complete", "status": "success", "tpu_found": true, "message": "Native Coral TPU skill installed -- Edge TPU ready"}
+    exit /b 0
+) else (
+    echo {"event": "complete", "status": "partial", "tpu_found": false, "message": "Native Coral TPU skill installed -- no TPU detected (CPU fallback)"}
+    exit /b 2
+)