SafeClaw

Security-First Private AI Assistant - Telegram Controlled

SafeClaw is a security-focused private AI assistant that operates through Telegram. Designed with privacy and security as primary concerns, SafeClaw features end-to-end encrypted conversations, secure credential management, and self-hosted deployment for complete control. GitHub: https://github.com/princezuda/safeclaw Developer: princezuda (Community Open Source) License: MIT Key Innovation: SafeClaw combines the convenience of Telegram's messaging platform with robust security features, creating a private AI assistant you can access from anywhere.

Core Philosophy: "Security first, convenience second, but never sacrifice privacy"

Why SafeClaw?

Telegram Interface: Access your AI assistant from any device with Telegram installed.

Encrypted Conversations: Leverage Telegram's encryption for secure AI interactions.

Self-Hosted: Full control over your data - nothing leaves your server except encrypted messages.

Credential Vault: Securely store and use API keys, passwords, and sensitive data.

Key Features

$#1.
Telegram Bot Interface

• Instant messaging interface
• Rich text formatting support
• File and media handling
• Voice message support
• Inline query support

$#2.
Security Features

• End-to-end encryption (via Telegram)
• Secure credential storage (encrypted at rest)
• Access control lists
• Session management
• Automatic logout

$#3.
Privacy Controls

• Local data processing
• No telemetry or analytics
• Configurable data retention
• Secure log handling
• Privacy-focused defaults

$#4.
Multi-User Support

• User isolation
• Per-user quotas
• Role-based access control
• Admin dashboard
• Usage statistics

$#5.
Audit and Compliance

• Complete audit logging
• Command history
• Access logs
• Export capabilities
• Compliance reporting

Installation

Docker (Recommended)

# Pull image
docker pull princezuda/safeclaw:latest

# Create config directory
mkdir -p ~/safeclaw/config
mkdir -p ~/safeclaw/data

# Create docker-compose.yml
docker-compose up -d

# Or run directly
docker run -d \
  --name safeclaw \
  -v ~/safeclaw/config:/app/config \
  -v ~/safeclaw/data:/app/data \
  -e TELEGRAM_BOT_TOKEN=your_token \
  -e ENCRYPTION_KEY=your_key \
  princezuda/safeclaw:latest

Manual Installation

# Clone repository
git clone https://github.com/princezuda/safeclaw.git
cd safeclaw

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Linux/macOS
# or
venv\Scripts\activate  # Windows

# Install dependencies
pip install -r requirements.txt

# Configure
cp config.example.yaml config.yaml
# Edit config.yaml

# Run
python main.py

Configuration

config.yaml

telegram:
  bot_token: ${TELEGRAM_BOT_TOKEN}
  allowed_users:
    - 123456789  # Your Telegram user ID

security:
  encryption_key: ${ENCRYPTION_KEY}
  session_timeout: 3600  # 1 hour
  max_login_attempts: 5

ai:
  provider: openai
  model: gpt-4o-mini
  # api_key stored in vault

storage:
  data_path: ./data
  log_path: ./logs
  retention_days: 30

features:
  file_processing: true
  voice_messages: true
  code_execution: false  # Disabled by default for security

Environment Variables

TELEGRAM_BOT_TOKEN=1234567890:ABCdefGHIjklMNOpqrsTUVwxyz
ENCRYPTION_KEY=your-32-character-secret-key-here

Pricing

Free: SafeClaw is completely free and open source under MIT license. You only pay for:

• Server hosting (if not running locally)
• AI model API usage
• Telegram is free for personal use

System Requirements

Use Cases

Secure Personal AI Assistant

Access your AI assistant securely from anywhere via Telegram, with full encryption.

Private Task Management

Manage tasks, reminders, and notes through encrypted Telegram messages.

Encrypted AI Conversations

Have sensitive conversations with AI, knowing they're encrypted end-to-end.

Remote Secure Automation

Trigger automations and workflows remotely through secure Telegram commands.

Team AI Assistant

Share an AI assistant with your team while maintaining access control and isolation.

Privacy-Sensitive Workflows

Run AI workflows that handle sensitive data without exposing it to third parties.

Security Best Practices

$#1.
Strong Encryption Key

# Generate a strong encryption key
openssl rand -hex 32

$#2.
Restrict Access

security:
  allowed_users:
    - 123456789  # Only your user ID
  admin_users:
    - 123456789

$#3.
Enable Audit Logging

audit:
  enabled: true
  log_path: ./secure-logs
  retention_days: 90

$#4.
Regular Updates

# Keep SafeClaw updated
docker pull princezuda/safeclaw:latest
docker-compose restart

Community and Support

• GitHub: https://github.com/princezuda/safeclaw
• Documentation: https://github.com/princezuda/safeclaw#readme
• Issues: GitHub Issues tab
• Security Reports: [email protected] (encrypted)

Quick Start Guide

Get SafeClaw up and running in about 10 minutes.

Option A: Docker (Recommended)

docker pull safeclaw/safeclaw:latest
docker run -d --name safeclaw safeclaw/safeclaw:latest

Option B: Manual Installation

git clone .git
cd safeclaw
pip install -r requirements.txt
python setup.py

Step 3: Configure Your AI Model

Set your preferred AI model and API key in the configuration file.

Step 4: Connect a Platform

Link your messaging platform (Telegram, Discord, etc.) and start sending commands.

Time to first result: ~10 minutes with Docker, ~20 minutes with manual setup.

Full documentation: https://github.com/princezuda/safeclaw#readme

FAQ

Is SafeClaw free to use?

Yes, SafeClaw is free and open source (MIT license). You only pay for AI model API costs if using external models.

What are the system requirements for SafeClaw?

SafeClaw requires 512MB RAM of RAM minimum. Runtime: Python/Docker. It runs on Windows, macOS, and Linux.

Can I self-host SafeClaw?

Yes. SafeClaw is open source (MIT) and can be self-hosted on your own hardware. Clone the repository from GitHub and follow the installation guide.

What platforms does SafeClaw support?

SafeClaw supports: Telegram. Check the official documentation for the most up-to-date list of supported platforms.

How does SafeClaw compare to OpenClaw?

SafeClaw offers a different approach compared to OpenClaw. While OpenClaw provides the largest ecosystem with 13,729+ skills and maximum flexibility, SafeClaw focuses on personal assistant. Choose SafeClaw if you prioritize its specific features; choose OpenClaw for the broadest compatibility and community support.

Is SafeClaw suitable for beginners?

SafeClaw requires some technical knowledge to set up (Python/Docker). If you are a beginner, consider starting with QClaw (one-click install) or MaxClaw (cloud-based, no setup) first, then graduate to SafeClaw as you gain experience.

License

MIT License - Free for personal and commercial use.

Tags

ai-agent, security, telegram, privacy, encryption, self-hosted, bot, secure