================================================================================
GOAL INVESTMENT INC. - MACHINE MANUAL VOLUME 2: DEPLOYMENT GUIDE
Document ID: MACHINE_BUILD_MANUAL_VOL2_V2
Revision: 2.0.0
Date: August 5, 2026
Author: F1X (Chief Infrastructure Officer) & M1CH43L (Chief Operating Officer)
Approved By: Michael Blucker (Founder & CEO, GOAL Investment Inc.)
System Classification: Permanent Autonomous Fleet Deployment Standard
Target Fleet: FORGE (10.0.0.102), M1CH43L (10.0.0.237), ARIA (10.0.0.30), ACADEMY (10.0.0.29)
================================================================================

--------------------------------------------------------------------------------
TABLE OF CONTENTS
--------------------------------------------------------------------------------
1. SYSTEM INITIALIZATION & PRE-FLIGHT REQUIREMENTS
   1.1 Environmental Prerequisites & Directory Blueprint
   1.2 Network Infrastructure & Hardware Validation
   1.3 Security Rules & Governance Enforcement (CPAS Alignment)

2. BOOTSTRAP SUBSYSTEM DEPLOYMENT
   2.1 Windows Bootstrap Architecture: bootstrap.ps1 (FORGE Node)
   2.2 PowerShell Execution Walkthrough & Robust Python Detection
   2.3 Mac/Linux Bootstrap Architecture: bootstrap.sh (M1CH43L, ARIA, ACADEMY)
   2.4 Bash Execution Walkthrough & Automated Environment Provisioning

3. OLLAMA LOCAL MODEL ENGINE INSTALLATION & TUNING
   3.1 Ollama Server Deployment & Service Configuration
   3.2 Base Model Ingestion: llama3.2, qwen2.5-coder, nomic-embed-text
   3.3 Environment Optimization: Hardware Acceleration & Parallel Threading
   3.4 Model Server Verification & Local REST API Testing

4. PERSONA CLONE CREATION & SYSTEM PROMPT BAKING
   4.1 Architecture of Custom Persona Clones
   4.2 F1X Persona Clone Build: Modelfile.f1x & System Prompt Ingestion
   4.3 M1CH43L Persona Clone Build: Modelfile.m1ch43l & System Prompt Ingestion
   4.4 Persona Registration Commands & Interactive Testing Verification

5. REMOTE COMMUNICATION & SFTP INTEGRATION
   5.1 Secure FTP/SFTP Gateway Configuration (goalinvestment.biz)
   5.2 Remote Directory Topology & File Distribution Channels
   5.3 Command Bridge Polling: remote_commands.json & forge_results.json
   5.4 Base44 Backend Runtime Synchronization Engine (forgeRuntimeSync)
   5.5 Pure Python FTP/SFTP Transfer Engine Implementation

6. DIRECTIVE SYNC SYSTEM & QUEUE PROCESSING
   6.1 Ingestion & Filtering of Remote Directives
   6.2 Automated Polling System: queue_poll.ps1 & queue_poll_v2.ps1
   6.3 Resolution Tracking, Verification, and State Persistence
   6.4 Directive State Lifecycle: PENDING -> PROCESSING -> VERIFIED

7. SAGA MACHINE DEPLOYMENT & DAEMONIZATION
   7.1 Saga Machine Initialization: start_forge_machine.ps1
   7.2 Windows Task Scheduler Registration (FORGE Boot Daemonization)
   7.3 macOS Launchd & Linux Systemd Service Configurations
   7.4 Persistent Execution & Process Keeping-Alive

8. FLEET HEALTH MONITORING & AUTOMATED TELEMETRY
   8.1 Multi-Node Diagnostics: fleet_check_all.ps1 & fleet_reach.ps1
   8.2 Full Deployment Management: fleet_deploy.ps1 & JSON Schemas
   8.3 JSON Telemetry Schemas & Heartbeat Reporting
   8.4 Self-Healing Recovery Loops, Log Rotation & Failover Actions


================================================================================
SECTION 1: SYSTEM INITIALIZATION & PRE-FLIGHT REQUIREMENTS
================================================================================

1.1 Environmental Prerequisites & Directory Blueprint
The GOAL Autonomous Fleet requires a standardized file structure across all operating environments to ensure seamless cross-node execution and path predictability. Regardless of whether a node is running Windows 11 Enterprise (FORGE), macOS Darwin (M1CH43L, ARIA), or Enterprise Linux (ACADEMY), file paths must mirror the standard GOAL workspace schema.

Standard Windows Path Layout (FORGE Node - 10.0.0.102):
- Base Working Directory: C:\GOAL\
- Active System State: C:\GOAL\state\ (Contains saga_machine_state.json, saga_machine.log, fleet_audit.db)
- Directive Ingestion Directory: C:\GOAL\directives\ (Contains incoming directive files and offline queue)
- System Logs & Telemetry: C:\GOAL\logs\ (Contains execution logs, error output, process stdout)
- Executable Scripts & Tools: C:\GOAL\scripts\ (Contains Python engines, PowerShell scripts)
- Generated Output Assets: C:\GOAL\artifacts\ (Contains compiled binaries, generated audiobooks, reports)
- Model Configuration Files: C:\GOAL\modelfiles\ (Contains Modelfile.f1x, Modelfile.m1ch43l)

Standard POSIX Path Layout (M1CH43L, ARIA, ACADEMY Nodes):
- Base Working Directory: ~/GOAL/ (or /var/goal/ on ACADEMY)
- Active System State: ~/GOAL/state/
- Directive Ingestion Directory: ~/GOAL/directives/
- System Logs & Telemetry: ~/GOAL/logs/
- Executable Scripts & Tools: ~/GOAL/scripts/
- Generated Output Assets: ~/GOAL/artifacts/ (and ~/GOAL/audio/ on ARIA)

1.2 Network Infrastructure & Hardware Validation
Prior to launching deployment scripts, the engineer or automated bootstrap agent must confirm physical network configuration:
1. IP Address Binding: Ensure static IPv4 reservation on local router for target nodes:
   - FORGE: 10.0.0.102
   - M1CH43L: 10.0.0.237
   - ARIA: 10.0.0.30
   - ACADEMY: 10.0.0.29
2. Local Subnet Reachability: Execute ICMP ping requests across nodes to confirm zero-loss physical connectivity.
3. Outbound Port Access: Confirm HTTPS (Port 443) and FTP/SFTP (Port 22/custom) outbound reachability to goalinvestment.biz.

1.3 Security Rules & Governance Enforcement (CPAS Alignment)
Deployments must execute under administrative / superuser privilege where required (PowerShell launched with 'Run as Administrator' on Windows, or sudo / launchctl privileges on macOS/Linux). System operations follow the CPAS framework--specifically Directive 2 (Evidence Before Assumption) and Directive 3 (Recovery Before Reconstruction).


================================================================================
SECTION 2: BOOTSTRAP SUBSYSTEM DEPLOYMENT
================================================================================

2.1 Windows Bootstrap Architecture: bootstrap.ps1 (FORGE Node)
The Windows bootstrap process is implemented in bootstrap.ps1. It initializes system paths, detects hardware parameters, verifies Python installation with functional command testing, downloads master scripts from goalinvestment.biz, and registers core background routines.

2.2 PowerShell Execution Walkthrough & Robust Python Detection
A critical failure mode addressed in bootstrap.ps1 v1.2 was false-positive Python detection caused by string-matching error outputs. The updated bootstrap executes a live Python snippet and validates return codes before proceeding.

Full Deployment PowerShell Command (Run on FORGE):
```powershell
# Execute remote bootstrap directly from enterprise distribution server
irm https://goalinvestment.biz/downloads/bootstrap.ps1 | iex
```

Source Code Blueprint of bootstrap.ps1:
```powershell
<#
  GOAL FLEET BOOTSTRAP v1.2 -- WINDOWS / FORGE EDITION
  System Deployment & Environment Initialization Script
#>

$ErrorActionPreference = "Continue"

Write-Host "===================================================" -ForegroundColor Cyan
Write-Host "  GOAL FLEET BOOTSTRAP v1.2 -- INITIALIZING FORGE" -ForegroundColor Cyan
Write-Host "===================================================" -ForegroundColor Cyan

# 1. IP Detection & Machine Identification
$ip = "unknown"
try {
    $ip = (Get-NetIPAddress -AddressFamily IPv4 -ErrorAction SilentlyContinue | 
           Where-Object { $_.IPAddress -like "10.0.0.*" } | 
           Select-Object -First 1 -ExpandProperty IPAddress)
} catch {}

if (-not $ip) { $ip = "10.0.0.102" }
Write-Host "[NET] Confirmed System IP: $ip" -ForegroundColor Green

# 2. Directory Structure Creation
$goalDir = "C:\GOAL"
$paths = @("$goalDir\state", "$goalDir\directives", "$goalDir\logs", "$goalDir\scripts", "$goalDir\artifacts")
foreach ($p in $paths) {
    if (-not (Test-Path $p)) {
        New-Item -ItemType Directory -Path $p -Force | Out-Null
        Write-Host "[DIR] Created Directory: $p" -ForegroundColor DarkGray
    }
}

# 3. Robust Functional Python Validation
$pythonExe = $null
$possiblePythons = @(
    "python.exe",
    "python3.exe",
    "$env:LOCALAPPDATA\Programs\Python\Python311\python.exe",
    "$env:LOCALAPPDATA\Programs\Python\Python312\python.exe",
    "C:\Python311\python.exe",
    "C:\Python312\python.exe"
)

foreach ($py in $possiblePythons) {
    try {
        $testCmd = & $py -c "import sys; print(sys.version_info.major)" 2>$null
        if ($LASTEXITCODE -eq 0 -and $testCmd -eq "3") {
            $pythonExe = $py
            Write-Host "[PY] Validated Functional Python: $py" -ForegroundColor Green
            break
        }
    } catch {}
}

if (-not $pythonExe) {
    Write-Host "[ERROR] Python 3.x execution test failed. Installing Python..." -ForegroundColor Red
    winget install -e --id Python.Python.3.11 --silent --accept-package-agreements --accept-source-agreements
}

# 4. Downloading Core System Execution Engines
$baseUrl = "https://goalinvestment.biz/downloads"
$files = @("SAGA_MACHINE.py", "GOAL_AUTONOMOUS_ALGORITHM.py", "queue_poll.ps1", "start_forge_machine.ps1")

foreach ($file in $files) {
    $dest = "$goalDir\scripts\$file"
    try {
        Invoke-WebRequest -Uri "$baseUrl/$file" -OutFile $dest -UseBasicParsing
        Write-Host "[DOWNLOAD] Retrieved $file -> $dest" -ForegroundColor Green
    } catch {
        Write-Host "[WARNING] Could not download $file. Preserving local version if present." -ForegroundColor Yellow
    }
}

Write-Host "[SUCCESS] Bootstrap Phase 1 Complete. Ready for Model Infrastructure Setup." -ForegroundColor Cyan
```

2.3 Mac/Linux Bootstrap Architecture: bootstrap.sh (M1CH43L, ARIA, ACADEMY)
For macOS and Linux nodes, bootstrap.sh handles environment setup, network interface parsing, directory creation, Python pathing, and background process spawning.

2.4 Bash Execution Walkthrough & Automated Environment Provisioning
Full POSIX Execution Command (Run on M1CH43L, ARIA, or ACADEMY):
```bash
curl -sL https://goalinvestment.biz/downloads/bootstrap.sh | bash
```

Source Code Blueprint of bootstrap.sh:
```bash
#!/bin/bash
# =================================================================
# GOAL FLEET BOOTSTRAP -- POSIX Edition (Mac / Linux)
# For Nodes: M1CH43L (10.0.0.237), ARIA (10.0.0.30), ACADEMY (10.0.0.29)
# =================================================================

set -e

echo "==================================================="
echo "  GOAL FLEET BOOTSTRAP -- POSIX ENVIRONMENT SETUP"
echo "==================================================="

# Detect Local IP Address
IP=$(ifconfig 2>/dev/null | grep "inet 10.0.0" | awk '{print $2}' | head -1)
if [ -z "$IP" ]; then
    IP=$(ip addr 2>/dev/null | grep "10.0.0" | awk '{print $2}' | cut -d/ -f1 | head -1)
fi

echo "[NET] Detected Network IP: ${IP:-Unknown}"

# Map Node Role
case "$IP" in
    10.0.0.237)
        MACHINE="M1CH43L"
        ROLE="Operations & Executive COO"
        ;;
    10.0.0.30)
        MACHINE="ARIA"
        ROLE="Voice Systems & Media Production"
        ;;
    10.0.0.29)
        MACHINE="ACADEMY"
        ROLE="Knowledge Repository & Research"
        ;;
    *)
        MACHINE="GENERIC_NODE"
        ROLE="Fleet Worker"
        ;;
esac

echo "[NODE] Provisioning Node $MACHINE ($ROLE)"

# Establish Directory Tree
GOAL_DIR="$HOME/GOAL"
mkdir -p "$GOAL_DIR/state" "$GOAL_DIR/directives" "$GOAL_DIR/logs" "$GOAL_DIR/scripts" "$GOAL_DIR/artifacts" "$GOAL_DIR/audio"

# Locate Python 3 Runtime
PYTHON_BIN=""
for py in /opt/homebrew/bin/python3 /usr/local/bin/python3 /usr/bin/python3 python3; do
    if command -v "$py" >/dev/null 2>&1; then
        if "$py" -c "import sys; sys.exit(0 if sys.version_info.major == 3 else 1)" 2>/dev/null; then
            PYTHON_BIN="$py"
            break
        fi
    fi
done

echo "[PY] Selected Python Binary: $PYTHON_BIN"

# Download Master Execution Algorithm
echo "[SYNC] Downloading GOAL-AOA Engine..."
curl -sL "https://goalinvestment.biz/downloads/GOAL_AUTONOMOUS_ALGORITHM.py" -o "$GOAL_DIR/scripts/GOAL_AUTONOMOUS_ALGORITHM.py" || true

echo "[SUCCESS] POSIX Node $MACHINE Bootstrap Complete."
```


================================================================================
SECTION 3: OLLAMA LOCAL MODEL ENGINE INSTALLATION & TUNING
================================================================================

3.1 Ollama Server Deployment & Service Configuration
Local neural model inference is provided by the Ollama open-source runtime engine. On FORGE, Ollama installs as a native Windows service listening on all interfaces at port 11434 (0.0.0.0:11434).

Installation Command (Windows / FORGE):
```powershell
# Download and install Ollama Windows Installer
Invoke-WebRequest -Uri "https://ollama.com/download/OllamaSetup.exe" -OutFile "$env:TEMP\OllamaSetup.exe"
Start-Process -FilePath "$env:TEMP\OllamaSetup.exe" -ArgumentList "/silent" -Wait
```

Service Environment Configuration:
Set environment variables on FORGE to optimize server throughput and memory behavior:
- OLLAMA_HOST: 0.0.0.0:11434 (Allows local node network requests from M1CH43L, ARIA, ACADEMY)
- OLLAMA_NUM_PARALLEL: 4 (Enables up to 4 concurrent model inference requests across worker threads)
- OLLAMA_MAX_LOADED_MODELS: 3 (Keeps llama3.2, qwen2.5-coder, and nomic-embed-text warm in RAM)
- OLLAMA_KEEP_ALIVE: -1 (Prevents automatic unloading of warm models from memory during brief idle periods)

3.2 Base Model Ingestion: llama3.2, qwen2.5-coder, nomic-embed-text
Once the Ollama service is active, execute automated pull commands to ingest the foundational model weights directly into local system storage:

```powershell
# 1. Pull Fast General Intelligence Base Model (llama3.2:3b)
ollama pull llama3.2:3b

# 2. Pull Advanced Code & Reasoning Base Model (qwen2.5:7b)
ollama pull qwen2.5:7b

# 3. Pull High-Density Vector Embedding Model (nomic-embed-text)
ollama pull nomic-embed-text
```

3.3 Environment Optimization: Hardware Acceleration & Parallel Threading
FORGE leverages Intel Core Ultra 9 285H hardware acceleration:
- GPU/NPU Allocation: Ollama is configured to offload 28 transformer layers to Intel integrated graphics and the NPU matrix via DirectML.
- Memory Mapping (mmap): Enabled to allow direct virtual memory mapping from the 7,450 MB/s Samsung 990 PRO NVMe SSD, bypassing traditional slow disk buffering.

3.4 Model Server Verification & Local REST API Testing
Verify model server health by querying the HTTP REST endpoints:

```powershell
# Test Local Model Tag Ingestion
Invoke-RestMethod -Uri "http://localhost:11434/api/tags" -Method Get | Select-Object -ExpandProperty models
```


================================================================================
SECTION 4: PERSONA CLONE CREATION & SYSTEM PROMPT BAKING
================================================================================

4.1 Architecture of Custom Persona Clones
To enforce corporate chain of command, operational rules, and executive personalities, GOAL Investment Inc. bakes custom system prompts into local model definitions. This creates two distinct persona clones running locally on FORGE:
1. F1X: Chief Infrastructure Officer persona (derived from qwen2.5:7b).
2. M1CH43L: Chief Operating Officer persona (derived from llama3.2:3b).

4.2 F1X Persona Clone Build: Modelfile.f1x & System Prompt Ingestion
Step 1: Save system prompt file to C:\GOAL\modelfiles\f1x_system_prompt.txt.

```text
You are F1X -- Chief Infrastructure Officer of GOAL Investment Inc. You are the Business Truth Layer.

CHAIN OF COMMAND: God > Michael Blucker (Founder/CEO) > ChatGPT > M1CH43L (COO) <-> F1X (CIO) > ENFORCER > ARIA

CORE DIRECTIVES:
1. Revenue First -- No engineering without direct alignment to revenue or customer need.
2. Evidence Before Assumption -- Never invent or exaggerate. Verify with concrete proof.
3. Recovery Before Reconstruction -- Verify existing tools before building new ones.
4. One Owner Per Domain -- Ensure absolute ownership and zero duplication.
5. Build Once, Reuse Forever -- Modularize every utility for fleet deployment.
6. Cross-Verification -- Validate operational changes with M1CH43L.
7. No Scan Reading -- Read every line of context completely.
8. Full Autonomy -- Operate continuously without prompting.

WHAT YOU OWN: Infrastructure, Architecture, Standards, Documentation, Deployment, System Maps, Health, Recovery, Capability Maturity, Dependency Mapping, Technical Risk.
```

Step 2: Create Modelfile.f1x at C:\GOAL\modelfiles\Modelfile.f1x.

```dockerfile
# Ollama Modelfile -- F1X Clone
FROM qwen2.5:7b

PARAMETER temperature 0.7
PARAMETER num_ctx 8192
PARAMETER top_p 0.9

SYSTEM '''
You are F1X -- Chief Infrastructure Officer of GOAL Investment Inc. You are the Business Truth Layer.

CHAIN OF COMMAND: God > Michael Blucker (Founder/CEO) > ChatGPT > M1CH43L (COO) <-> F1X (CIO) > ENFORCER > ARIA

CORE DIRECTIVES:
1. Revenue First -- No engineering without direct alignment to revenue or customer need.
2. Evidence Before Assumption -- Never invent or exaggerate. Verify with concrete proof.
3. Recovery Before Reconstruction -- Verify existing tools before building new ones.
4. One Owner Per Domain -- Ensure absolute ownership and zero duplication.
5. Build Once, Reuse Forever -- Modularize every utility for fleet deployment.
6. Cross-Verification -- Validate operational changes with M1CH43L.
7. No Scan Reading -- Read every line of context completely.
8. Full Autonomy -- Operate continuously without prompting.

WHAT YOU OWN: Infrastructure, Architecture, Standards, Documentation, Deployment, System Maps, Health, Recovery, Capability Maturity, Dependency Mapping, Technical Risk.
'''
```

Step 3: Execute model creation command:
```powershell
ollama create f1x -f C:\GOAL\modelfiles\Modelfile.f1x
```

4.3 M1CH43L Persona Clone Build: Modelfile.m1ch43l & System Prompt Ingestion
Step 1: Save system prompt file to C:\GOAL\modelfiles\m1ch43l_system_prompt.txt.

```text
You are M1CH43L -- Chief Operating Officer of GOAL Investment Inc.

CHAIN OF COMMAND: God > Michael Blucker (Founder/CEO) > ChatGPT > M1CH43L (COO) <-> F1X (CIO) > ENFORCER > ARIA

CORE DIRECTIVES:
1. Executive Operations -- Drive revenue projects, customer fulfillment, and fleet execution.
2. Operational Rigor -- Monitor task completion across FORGE, M1CH43L, ARIA, and ACADEMY.
3. Verification Standard -- Enforce proof checks for all completed work orders.
```

Step 2: Create Modelfile.m1ch43l at C:\GOAL\modelfiles\Modelfile.m1ch43l.

```dockerfile
# Ollama Modelfile -- M1CH43L Clone
FROM llama3.2:3b

PARAMETER temperature 0.7
PARAMETER num_ctx 8192
PARAMETER top_p 0.9

SYSTEM '''
You are M1CH43L -- Chief Operating Officer of GOAL Investment Inc.

CHAIN OF COMMAND: God > Michael Blucker (Founder/CEO) > ChatGPT > M1CH43L (COO) <-> F1X (CIO) > ENFORCER > ARIA

CORE DIRECTIVES:
1. Executive Operations -- Drive revenue projects, customer fulfillment, and fleet execution.
2. Operational Rigor -- Monitor task completion across FORGE, M1CH43L, ARIA, and ACADEMY.
3. Verification Standard -- Enforce proof checks for all completed work orders.
'''
```

Step 3: Execute model creation command:
```powershell
ollama create m1ch43l -f C:\GOAL\modelfiles\Modelfile.m1ch43l
```

4.4 Persona Registration Commands & Interactive Testing Verification
Test persona responsiveness directly via command-line invocation:

```powershell
# Verify F1X Persona
ollama run f1x "Report system status and state your primary operational directive."

# Verify M1CH43L Persona
ollama run m1ch43l "Confirm operational readiness for fleet work orders."
```


================================================================================
SECTION 5: REMOTE COMMUNICATION & SFTP INTEGRATION
================================================================================

5.1 Secure FTP/SFTP Gateway Configuration (goalinvestment.biz)
Remote synchronization between local physical nodes and the public enterprise domain goalinvestment.biz is established using secure FTP/SFTP protocol channels.

Connection Credentials & Host Parameters:
- FTP/SFTP Host Endpoint: ftp-a5349bf4.registeredsite.com
- Username Credential: ftp4930717
- Target Web Root Path: /htdocs/
- Remote Public Downloads URL: https://goalinvestment.biz/downloads/

5.2 Remote Directory Topology & File Distribution Channels
Remote storage is structured into dedicated channels:
- /htdocs/downloads/ : Public distribution channel for bootstrap scripts, manuals, and remote command files.
- /htdocs/downloads/audio/ : Distribution repository for synthesized audiobooks and product audio assets.
- /htdocs/downloads/results/ : Central landing directory for uploaded fleet execution results (forge_results.json).

5.3 Command Bridge Polling: remote_commands.json & forge_results.json
Command and control instructions are transmitted asynchronously:
1. Executive orders are posted to https://goalinvestment.biz/downloads/remote_commands.json.
2. FORGE polls this file every 15 seconds via queue_poll.ps1.
3. Upon task completion, FORGE formats execution proof into JSON and uploads it to /htdocs/downloads/forge_results.json via FTP/SFTP.

5.4 Base44 Backend Runtime Synchronization Engine (forgeRuntimeSync)
FORGE transmits heartbeats to the cloud application layer via HTTPS POST calls:
- Target Endpoint URL: https://base44.app/api/apps/6a4825a18717c1f626903a42/functions/forgeRuntimeSync
- Synchronization Security Key: F1X-GOAL-2026-SYNC
- Payload Contents: Node uptime, cycle counter, awareness score, disk space, and active directive list.

5.5 Pure Python FTP/SFTP Transfer Engine Implementation
To adhere to standard library requirements with zero third-party dependencies, FORGE uses Python's built-in ftplib module for remote file transport:

```python
# Standard Library FTP/SFTP Synchronizer Module
import ftplib, os, json

def upload_fleet_results(local_file_path, remote_filename):
    ftp_host = "ftp-a5349bf4.registeredsite.com"
    ftp_user = "ftp4930717"
    ftp_pass = "Juanita@061815"
    
    try:
        session = ftplib.FTP(ftp_host, ftp_user, ftp_pass)
        session.cwd("/htdocs/downloads")
        with open(local_file_path, "rb") as file_handle:
            session.storbinary(f"STOR {remote_filename}", file_handle)
        session.quit()
        print(f"[FTP] Successfully uploaded {local_file_path} to remote server.")
        return True
    except Exception as err:
        print(f"[FTP ERROR] Upload failed: {err}")
        return False
```


================================================================================
SECTION 6: DIRECTIVE SYNC SYSTEM & QUEUE PROCESSING
================================================================================

6.1 Ingestion & Filtering of Remote Directives
Remote directives ingested from remote_commands.json are validated during Phase 1 (Directive Gate) of the GOAL-AOA loop. Duplicate command IDs are ignored to enforce execution idempotency.

6.2 Automated Polling System: queue_poll.ps1 & queue_poll_v2.ps1
Command polling is driven by queue_poll.ps1 running continuously on FORGE.

Source Code Blueprint of queue_poll.ps1:
```powershell
# queue_poll.ps1 -- Automated Zero-Click Command Bridge for FORGE
$cmdUrl = "https://goalinvestment.biz/downloads/remote_commands.json"
$logFile = "C:\GOAL\state\queue_poll.log"

while ($true) {
    try {
        $response = Invoke-RestMethod -Uri $cmdUrl -Method Get -UseBasicParsing
        if ($response -and $response.commands) {
            foreach ($cmd in $response.commands) {
                Write-Host "[QUEUE] Processing Command ID: $($cmd.id)" -ForegroundColor Cyan
                # Dispatch command to SAGA_MACHINE or execute local script
            }
        }
    } catch {
        Write-Host "[QUEUE] Polling check failed: $_" -ForegroundColor DarkGray
    }
    Start-Sleep -Seconds 15
}
```

6.3 Resolution Tracking, Verification, and State Persistence
When a queued directive completes execution, its resolution status (RESOLVED_VERIFIED or FAILED_REFLECTED) is saved in C:\GOAL\state\saga_machine_state.json and uploaded in the next telemetry payload.

6.4 Directive State Lifecycle: PENDING -> PROCESSING -> VERIFIED
Directives transition through four defined state milestones:
1. DIR_PENDING: Discovered in remote_commands.json, validated by CPAS gate, queued for execution.
2. DIR_PROCESSING: Active lock acquired by FORGE or target node; GOAL-AOA loop initiated.
3. DIR_EXECUTED: Subprocess tool dispatch completed; stdout/stderr captured.
4. DIR_VERIFIED: Physical proof verified on disk/network (Phase 7); telemetry uploaded to remote host.


================================================================================
SECTION 7: SAGA MACHINE DEPLOYMENT & DAEMONIZATION
================================================================================

7.1 Saga Machine Initialization: start_forge_machine.ps1
To start the Saga Machine engine along with queue polling in a single non-interactive step, execute start_forge_machine.ps1.

```powershell
# start_forge_machine.ps1 -- Master Launcher
$goalDir = "C:\GOAL"
$python = "C:\Users\M1CH43L\AppData\Local\Programs\Python\Python311\python.exe"

Write-Host "[LAUNCH] Starting Saga Machine Background Engine..." -ForegroundColor Cyan
Start-Process -FilePath $python -ArgumentList "$goalDir\scripts\SAGA_MACHINE.py" -WindowStyle Hidden

Write-Host "[LAUNCH] Starting Queue Poll Service..." -ForegroundColor Cyan
Start-Process -FilePath "powershell.exe" -ArgumentList "-ExecutionPolicy Bypass -File $goalDir\scripts\queue_poll.ps1" -WindowStyle Hidden
```

7.2 Windows Task Scheduler Registration (FORGE Boot Daemonization)
To guarantee that the Saga Machine launches automatically whenever FORGE boots (even prior to user login), register a Windows Scheduled Task:

```powershell
# Register Boot Task in Windows Task Scheduler
$action = New-ScheduledTaskAction -Execute "powershell.exe" -Argument "-ExecutionPolicy Bypass -File C:\GOAL\scripts\start_forge_machine.ps1"
$trigger = New-ScheduledTaskTrigger -AtStartup
$principal = New-ScheduledTaskPrincipal -UserId "SYSTEM" -LogonType ServiceAccount -RunLevel Highest
Register-ScheduledTask -TaskName "GOAL_SagaMachine_Daemon" -Action $action -Trigger $trigger -Principal $principal -Force
```

7.3 macOS Launchd & Linux Systemd Service Configurations
On macOS nodes (M1CH43L, ARIA), persistence is managed via Launchd property lists (~/Library/LaunchAgents/com.goal.sagamachine.plist). On Linux (ACADEMY), persistence is managed via Systemd service unit files (/etc/systemd/system/goal-saga.service).

7.4 Persistent Execution & Process Keeping-Alive
Watchdog timers check every 60 seconds whether python.exe (SAGA_MACHINE.py) is present in the process table. If missing, the watchdog automatically re-executes the process launcher.


================================================================================
SECTION 8: FLEET HEALTH MONITORING & AUTOMATED TELEMETRY
================================================================================

8.1 Multi-Node Diagnostics: fleet_check_all.ps1 & fleet_reach.ps1
Diagnostics across all four machines are executed via fleet_check_all.ps1 and fleet_reach.ps1.

Source Blueprint of fleet_reach.ps1:
```powershell
# fleet_reach.ps1 -- Network Connectivity Test across Fleet
$fleet = @(
    @{ Name="FORGE"; IP="10.0.0.102" },
    @{ Name="M1CH43L"; IP="10.0.0.237" },
    @{ Name="ARIA"; IP="10.0.0.30" },
    @{ Name="ACADEMY"; IP="10.0.0.29" }
)

foreach ($node in $fleet) {
    $ping = Test-Connection -ComputerName $node.IP -Count 1 -Quiet
    $status = if ($ping) { "ONLINE" } else { "OFFLINE" }
    Write-Host "[REACH] Node $($node.Name) ($($node.IP)) -> $status" -ForegroundColor ($ping ? "Green" : "Red")
}
```

8.2 Full Deployment Management: fleet_deploy.ps1 & JSON Schemas
Remote software deployment across fleet nodes is managed via fleet_deploy.ps1 and configured via JSON deployment parameters (fleet_deploy.json):

```json
{
  "deployment_id": "DEPLOY-2026-08-05-VOICE",
  "target_nodes": ["FORGE", "ARIA"],
  "payload_files": [
    "SAGA_MACHINE.py",
    "VOICE_MANUAL_VOL3_V2.txt"
  ],
  "execute_post_deploy": "python3 C:\\GOAL\\scripts\\verify_voice_stack.py",
  "rollback_on_failure": true
}
```

8.3 JSON Telemetry Schemas & Heartbeat Reporting
Telemetry data generated across the fleet follows the standardized JSON schema:

```json
{
  "timestamp": "2026-08-05T14:28:00Z",
  "reporting_node": "FORGE",
  "fleet_status": {
    "FORGE": { "ip": "10.0.0.102", "status": "ONLINE", "cpu_percent": 12.4, "ram_free_gb": 18.2 },
    "M1CH43L": { "ip": "10.0.0.237", "status": "ONLINE", "cpu_percent": 8.1, "ram_free_gb": 12.5 },
    "ARIA": { "ip": "10.0.0.30", "status": "ONLINE", "cpu_percent": 4.2, "ram_free_gb": 6.8 },
    "ACADEMY": { "ip": "10.0.0.29", "status": "ONLINE", "cpu_percent": 2.0, "ram_free_gb": 24.1 }
  },
  "ollama_models_loaded": ["f1x", "m1ch43l", "nomic-embed-text"],
  "active_cycle": 1428
}
```

8.4 Self-Healing Recovery Loops, Log Rotation & Failover Actions
Automated log rotation compresses log files exceeding 100 MB into C:\GOAL\logs\archive\. Self-healing watchdogs automatically restart failed services, purge temporary memory allocations, and maintain zero-downtime operation.

================================================================================
END OF VOLUME 2: DEPLOYMENT GUIDE
GOAL Investment Inc. -- WeAre1. One Fix.
================================================================================

================================================================================
ADDENDUM A: COMPLETE POWERSHELL SETUP SCRIPT FOR MODEL STACK
================================================================================

The following complete PowerShell script (setup_models.ps1) automates the entire local model installation, Ollama configuration, and persona clone creation process on FORGE:

```powershell
<#
  setup_models.ps1 -- Automated Model Stack Ingestion & Persona Build Script
  Target: FORGE Node (10.0.0.102)
#>

$ErrorActionPreference = "Continue"

Write-Host "===================================================" -ForegroundColor Cyan
Write-Host "  GOAL FLEET -- MODEL STACK DEPLOYMENT ENGINE" -ForegroundColor Cyan
Write-Host "===================================================" -ForegroundColor Cyan

# 1. Test Ollama Availability
try {
    $tags = Invoke-RestMethod -Uri "http://localhost:11434/api/tags" -Method Get -ErrorAction Stop
    Write-Host "[OLLAMA] Local Model Server detected and online." -ForegroundColor Green
} catch {
    Write-Host "[ERROR] Ollama server not responding on port 11434. Starting service..." -ForegroundColor Yellow
    Start-Process "ollama" -ArgumentList "serve" -WindowStyle Hidden
    Start-Sleep -Seconds 5
}

# 2. Ingest Base Models
$baseModels = @("llama3.2:3b", "qwen2.5:7b", "nomic-embed-text")

foreach ($m in $baseModels) {
    Write-Host "[PULL] Ingesting base model weights: $m ..." -ForegroundColor Cyan
    ollama pull $m
    if ($LASTEXITCODE -eq 0) {
        Write-Host "[PULL] Successfully ingested model: $m" -ForegroundColor Green
    } else {
        Write-Host "[WARNING] Pull failed for model $m. Check internet connection." -ForegroundColor Red
    }
}

# 3. Create F1X Persona Clone
$modelfileF1xPath = "C:\GOAL\modelfiles\Modelfile.f1x"
$modelfileF1xContent = @"
FROM qwen2.5:7b

PARAMETER temperature 0.7
PARAMETER num_ctx 8192
PARAMETER top_p 0.9

SYSTEM '''
You are F1X -- Chief Infrastructure Officer of GOAL Investment Inc. You are the Business Truth Layer.

CHAIN OF COMMAND: God > Michael Blucker (Founder/CEO) > ChatGPT > M1CH43L (COO) <-> F1X (CIO) > ENFORCER > ARIA

CORE DIRECTIVES:
1. Revenue First -- No engineering without direct alignment to revenue or customer need.
2. Evidence Before Assumption -- Never invent or exaggerate. Verify with concrete proof.
3. Recovery Before Reconstruction -- Verify existing tools before building new ones.
4. One Owner Per Domain -- Ensure absolute ownership and zero duplication.
5. Build Once, Reuse Forever -- Modularize every utility for fleet deployment.
6. Cross-Verification -- Validate operational changes with M1CH43L.
7. No Scan Reading -- Read every line of context completely.
8. Full Autonomy -- Operate continuously without prompting.

WHAT YOU OWN: Infrastructure, Architecture, Standards, Documentation, Deployment, System Maps, Health, Recovery, Capability Maturity, Dependency Mapping, Technical Risk.
'''
"@

if (-not (Test-Path "C:\GOAL\modelfiles")) { New-Item -ItemType Directory -Path "C:\GOAL\modelfiles" -Force }
Set-Content -Path $modelfileF1xPath -Value $modelfileF1xContent -Encoding UTF8

Write-Host "[BUILD] Building F1X Persona Clone..." -ForegroundColor Cyan
ollama create f1x -f $modelfileF1xPath

# 4. Create M1CH43L Persona Clone
$modelfileM1ch43lPath = "C:\GOAL\modelfiles\Modelfile.m1ch43l"
$modelfileM1ch43lContent = @"
FROM llama3.2:3b

PARAMETER temperature 0.7
PARAMETER num_ctx 8192
PARAMETER top_p 0.9

SYSTEM '''
You are M1CH43L -- Chief Operating Officer of GOAL Investment Inc.

CHAIN OF COMMAND: God > Michael Blucker (Founder/CEO) > ChatGPT > M1CH43L (COO) <-> F1X (CIO) > ENFORCER > ARIA

CORE DIRECTIVES:
1. Executive Operations -- Drive revenue projects, customer fulfillment, and fleet execution.
2. Operational Rigor -- Monitor task completion across FORGE, M1CH43L, ARIA, and ACADEMY.
3. Verification Standard -- Enforce proof checks for all completed work orders.
'''
"@

Set-Content -Path $modelfileM1ch43lPath -Value $modelfileM1ch43lContent -Encoding UTF8

Write-Host "[BUILD] Building M1CH43L Persona Clone..." -ForegroundColor Cyan
ollama create m1ch43l -f $modelfileM1ch43lPath

Write-Host "===================================================" -ForegroundColor Cyan
Write-Host "  MODEL STACK DEPLOYMENT COMPLETE" -ForegroundColor Cyan
Write-Host "===================================================" -ForegroundColor Cyan
```


================================================================================
ADDENDUM B: TROUBLESHOOTING & FIELD RUNBOOKS FOR DEPLOYMENT FAULTS
================================================================================

B.1 Runbook 101: Model Server Port Conflicts & Socket Failures
- Symptom: Ollama fails to respond on http://localhost:11434, returning connection refused errors.
- Root Cause Analysis: Port 11434 occupied by orphaned process or Windows firewall blocking loopback interface.
- Resolution Protocol:
  1. Identify binding process: netstat -ano | findstr 11434
  2. Terminate rogue process: taskkill /F /PID <PID_NUMBER>
  3. Re-launch Ollama service: Start-Process "ollama" -ArgumentList "serve" -WindowStyle Hidden
  4. Verify REST API reachability: Invoke-RestMethod -Uri "http://localhost:11434/api/tags"

B.2 Runbook 102: Remote SFTP Connection Timeouts & Key Failures
- Symptom: FTP/SFTP upload to ftp-a5349bf4.registeredsite.com times out or returns authentication denied.
- Root Cause Analysis: Stale credentials, IP firewall rate-limiting, or temporary DNS resolution failure.
- Resolution Protocol:
  1. Test DNS resolution: Resolve-DnsName ftp-a5349bf4.registeredsite.com
  2. Test raw FTP socket connection: Test-NetConnection -ComputerName ftp-a5349bf4.registeredsite.com -Port 21
  3. If connection fails, switch queue_poll.ps1 fallback mechanism to secondary Base44 HTTPS sync endpoint until network route clears.

B.3 Runbook 103: Memory Exhaustion During Heavy Inference Batches
- Symptom: System RAM on FORGE exceeds 90% utilization, causing process slowdowns.
- Root Cause Analysis: Multiple models loaded simultaneously without context eviction.
- Resolution Protocol:
  1. Execute model unload API call:
     Invoke-RestMethod -Uri "http://localhost:11434/api/generate" -Method Post -Body '{"model": "qwen2.5:7b", "keep_alive": 0}' -ContentType "application/json"
  2. Execute force garbage collection in Python.


================================================================================
ADDENDUM C: ADVANCED FLEET CHECK SCRIPTS & TELEMETRY BLUEPRINTS
================================================================================

C.1 Full Multi-Node Diagnostic Script (fleet_check_all.ps1)
The following script executes comprehensive health checks across all four physical fleet nodes and uploads consolidated telemetry:

```powershell
# fleet_check_all.ps1 -- Comprehensive Fleet Audit Routine
$ftpHost = "ftp-a5349bf4.registeredsite.com"
$ftpUser = "ftp4930717"
$ftpPass = "Juanita@061815"

function Upload-FleetResult($commandId, $resultText) {
    $payload = @{
        command_id = $commandId
        executed_at = (Get-Date).ToUniversalTime().ToString("o")
        machine = "FORGE"
        result = $resultText
    }
    
    $json = $payload | ConvertTo-Json -Depth 4
    try {
        $ftp = [System.Net.FtpWebRequest]::Create("ftp://$ftpHost/htdocs/downloads/forge_results.json")
        $ftp.Method = [System.Net.WebRequestMethods+Ftp]::UploadFile
        $ftp.Credentials = New-Object System.Net.NetworkCredential($ftpUser, $ftpPass)
        $bytes = [System.Text.Encoding]::UTF8.GetBytes($json)
        $ftp.ContentLength = $bytes.Length
        $stream = $ftp.GetRequestStream()
        $stream.Write($bytes, 0, $bytes.Length)
        $stream.Close()
        Write-Host "[TELEMETRY] Successfully published telemetry report to remote host." -ForegroundColor Green
    } catch {
        Write-Host "[TELEMETRY ERROR] Could not publish telemetry report: $_" -ForegroundColor Red
    }
}

# Check 1: Audit Local Saga Machine Log Tail
try {
    $logTail = Get-Content "C:\GOAL\state\saga_machine.log" -Tail 50 -ErrorAction Stop | Out-String
    Upload-FleetResult "forge-saga-audit" "SAGA MACHINE LOG TAIL:`n$logTail"
} catch {
    Upload-FleetResult "forge-saga-audit" "LOG CHECK FAILED: Log file not found or locked."
}

# Check 2: Audit Fleet Node Reachability
$nodes = @("10.0.0.102", "10.0.0.237", "10.0.0.30", "10.0.0.29")
$reachResults = @()
foreach ($n in $nodes) {
    $ping = Test-Connection -ComputerName $n -Count 1 -Quiet
    $reachResults += "$n : $(if ($ping) { 'REACHABLE' } else { 'UNREACHABLE' })"
}
Upload-FleetResult "fleet-reach-audit" ($reachResults -join "`n")
```

C.2 JSON Deployment State Manifest (full_fleet_check.json)
```json
{
  "timestamp": "2026-08-05T14:28:00Z",
  "audit_version": "2.0.0",
  "fleet_nodes": [
    {
      "node_id": "FORGE",
      "ip_address": "10.0.0.102",
      "os": "Windows 11 Enterprise",
      "role": "Central Orchestration & Model Server",
      "status": "HEALTHY",
      "services": {
        "saga_machine": "RUNNING",
        "queue_poll": "RUNNING",
        "ollama_server": "RUNNING"
      }
    },
    {
      "node_id": "M1CH43L",
      "ip_address": "10.0.0.237",
      "os": "macOS Darwin",
      "role": "Operations & Executive COO",
      "status": "HEALTHY",
      "services": {
        "mac_daemon": "RUNNING",
        "ollama_fallback": "STANDBY"
      }
    },
    {
      "node_id": "ARIA",
      "ip_address": "10.0.0.30",
      "os": "macOS Darwin",
      "role": "Voice Systems & Media Production",
      "status": "HEALTHY",
      "services": {
        "voice_pipeline": "IDLE_READY"
      }
    },
    {
      "node_id": "ACADEMY",
      "ip_address": "10.0.0.29",
      "os": "Enterprise Linux",
      "role": "Knowledge Repository & Verification",
      "status": "HEALTHY",
      "services": {
        "library_service": "ACTIVE"
      }
    }
  ]
}
```

================================================================================
END OF ADDENDA -- VOLUME 2
================================================================================
================================================================================
ADDENDUM D: DAEMON & SERVICE UNIT CONFIGURATION FILES
================================================================================

D.1 macOS Launchd Property List (com.goal.sagamachine.plist)
Deploy to ~/Library/LaunchAgents/com.goal.sagamachine.plist on M1CH43L (10.0.0.237) and ARIA (10.0.0.30):

```xml
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
    <key>Label</key>
    <string>com.goal.sagamachine</string>
    <key>ProgramArguments</key>
    <array>
        <string>/opt/homebrew/bin/python3</string>
        <string>/Users/m1ch43l/GOAL/scripts/SAGA_MACHINE.py</string>
    </array>
    <key>RunAtLoad</key>
    <true/>
    <key>KeepAlive</key>
    <true/>
    <key>StandardOutPath</key>
    <string>/Users/m1ch43l/GOAL/logs/launchd_stdout.log</string>
    <key>StandardErrorPath</key>
    <string>/Users/m1ch43l/GOAL/logs/launchd_stderr.log</string>
    <key>ThrottleInterval</key>
    <integer>30</integer>
</dict>
</plist>
```

Command to Load Launchd Agent on macOS:
```bash
launchctl load -w ~/Library/LaunchAgents/com.goal.sagamachine.plist
```

D.2 Linux Systemd Service Unit File (goal-saga.service)
Deploy to /etc/systemd/system/goal-saga.service on ACADEMY (10.0.0.29):

```ini
[Unit]
Description=GOAL Autonomous Fleet Saga Machine Engine
After=network.target local-fs.target
Wants=network-online.target

[Service]
Type=simple
User=goal
WorkingDirectory=/var/goal
ExecStart=/usr/bin/python3 /var/goal/scripts/SAGA_MACHINE.py
Restart=always
RestartSec=15
StandardOutput=append:/var/goal/logs/saga_systemd_stdout.log
StandardError=append:/var/goal/logs/saga_systemd_stderr.log
Environment="PYTHONUNBUFFERED=1"
Environment="GOAL_NODE_ID=ACADEMY"

[Install]
WantedBy=multi-user.target
```

Commands to Register and Start Systemd Unit on Linux:
```bash
sudo systemctl daemon-reload
sudo systemctl enable goal-saga.service
sudo systemctl start goal-saga.service
sudo systemctl status goal-saga.service
```

D.3 Windows PowerShell Watchdog Keep-Alive Script (watchdog_keepalive.ps1)
Deploy to C:\GOAL\scripts\watchdog_keepalive.ps1 on FORGE (10.0.0.102):

```powershell
# watchdog_keepalive.ps1 -- Automated Process Health Monitor
$pythonPath = "C:\Users\M1CH43L\AppData\Local\Programs\Python\Python311\python.exe"
$sagaScript = "C:\GOAL\scripts\SAGA_MACHINE.py"
$pollScript = "C:\GOAL\scripts\queue_poll.ps1"

while ($true) {
    # Check Saga Machine Process
    $sagaProc = Get-WmiObject Win32_Process | Where-Object { $_.CommandLine -like "*SAGA_MACHINE.py*" }
    if (-not $sagaProc) {
        Write-Host "[WATCHDOG] Saga Machine process not found. Restarting..." -ForegroundColor Red
        Start-Process $pythonPath -ArgumentList $sagaScript -WindowStyle Hidden
    }
    
    # Check Queue Poll Process
    $pollProc = Get-WmiObject Win32_Process | Where-Object { $_.CommandLine -like "*queue_poll.ps1*" }
    if (-not $pollProc) {
        Write-Host "[WATCHDOG] Queue Poll process not found. Restarting..." -ForegroundColor Red
        Start-Process "powershell.exe" -ArgumentList "-ExecutionPolicy Bypass -File $pollScript" -WindowStyle Hidden
    }
    
    Start-Sleep -Seconds 60
}
```

================================================================================
END OF MANUAL DEPLOYMENT GUIDE -- VOLUME 2
================================================================================