| Framework Home | Examples Index | Cross-Pillar Governance | AISM | NEXUS | Dashboard |
Current framework context: AI SAFE² v3.1. This example may preserve historical component versions or earlier framework references where they describe when the implementation was created. For current conformance, use the v3.1 framework and applicable profile requirements.
Complete security toolkit for Ishi desktop agent users
New to Ishi security? → 10-Minute Hardening Guide
Having issues? → Troubleshooting Guide
Windows 11 setup? → Ishi + OpenClaw Integration
Quick Start: 5 Essential Integrations
Complete Guide: 30 Use Case Implementations
Categories:
File: ishi_memory.md
What it does: Persistent safety controls embedded in Ishi’s context memory.
Installation:
# Windows
cd $env:APPDATA\ishi\memories\
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/CyberStrategyInstitute/ai-safe2-framework/main/examples/ishi/ishi_memory.md" -OutFile "ishi_memory.md"
# Restart Ishi
# macOS/Linux
cd ~/.ishi/memories/
curl -O https://raw.githubusercontent.com/CyberStrategyInstitute/ai-safe2-framework/main/examples/ishi/ishi_memory.md
# Restart Ishi
Key protections:
File: ishi-scanner.py
What it does: Audits your Ishi installation for 10 security risks.
Installation:
# Windows
cd $env:USERPROFILE\Downloads
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/CyberStrategyInstitute/ai-safe2-framework/main/examples/ishi/ishi-scanner.py" -OutFile "ishi-scanner.py"
python ishi-scanner.py
# macOS/Linux
curl -O https://raw.githubusercontent.com/CyberStrategyInstitute/ai-safe2-framework/main/examples/ishi/ishi-scanner.py
python3 ishi-scanner.py
Checks:
Output: Risk score 0-100 + specific fixes
File: troubleshooting-ishi.md
20+ common issues with copy-paste fixes:
Platform-specific: Windows PowerShell scripts included
File: ishi-openclaw-integration.md
Architecture:
Time: 2 hours setup
Cost: $0/month (free tier works)
| File | Purpose | Time | Required? |
|---|---|---|---|
ishi_memory.md |
Safety protocol | 5 min | ⭐ Yes |
ishi-scanner.py |
Vulnerability scan | 2 min | ⭐ Yes |
troubleshooting-ishi.md |
Issue fixes | Reference | As needed |
ishi-openclaw-integration.md |
Full integration | 2 hours | Optional |
ai-safe2-framework/
├── examples/
│ ├── ishi/ # You are here
│ │ ├── README.md # This file
│ │ ├── ishi_memory.md # Memory protocol
│ │ ├── ishi-scanner.py # Security scanner
│ │ ├── troubleshooting-ishi.md # Issue resolution
│ │ └── ishi-openclaw-integration.md # Integration guide
│ └── openclaw/ # OpenClaw security
│ ├── README.md
│ ├── openclaw_memory.md
│ ├── scanner.py
│ ├── troubleshooting.md
│ └── gateway/ # Control gateway
└── guides/
├── ishi-hardening.md # 10-min setup
└── openclaw-hardening.md
Works with:
Token tracking prevents:
Auto-switches providers when limits hit
| Platform | Memory Protocol | Scanner | Notes |
|---|---|---|---|
| Windows 11 | ✅ | ✅ | PowerShell scripts included |
| macOS | ✅ | ✅ | Native support |
| Linux | ✅ | ✅ | Ubuntu/Debian tested |
Fix:
# Windows
Test-Path "$env:APPDATA\ishi\memories\ishi_memory.md"
# Should return: True
# If False, re-download
cd $env:APPDATA\ishi\memories\
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/CyberStrategyInstitute/ai-safe2-framework/main/examples/ishi/ishi_memory.md" -OutFile "ishi_memory.md"
Fix:
python ishi-scanner.py --path "$env:APPDATA\ishi"
Fix:
In Ishi: /config ghost_files
Set: Auto-commit = DISABLED
Full troubleshooting: troubleshooting-ishi.md
Documentation: AI SAFE² Framework
Issues: GitHub Issues
Security: Report vulnerability
Ishi Community: Discord
Start here: 10-Minute Hardening Guide
Built by: Cyber Strategy Institute
License: MIT (code) + CC-BY-SA 4.0 (documentation)
Version: 2.1
| Examples Index | Framework Home | Cross-Pillar Governance | NEXUS | Scanner | MCP Profile |
AI SAFE² v3.1 | Cyber Strategy Institute