Guardrail

Warning

Code in this repository is not audited and may contain serious security holes. Use at your own risk.

Guardrail is a security-focused guard contract that provides enhanced control over DELEGATECALL operations in Safe smart accounts. It implements a time-delayed allowlist system for delegate contracts, helping prevent unauthorized delegate calls while maintaining flexibility for legitimate use cases.

Overview

Guardrail acts as a security layer for Safe smart accounts by:

• Restricting delegate calls to pre-approved contracts only
• Implementing time delays for adding new delegates (configurable)
• Allowing immediate removal of delegates when needed
• Supporting both transaction and module transaction flows in Safe v1.5.0+

Key Features

• Immediate Delegate Addition: Add delegates instantly when the guard is not yet enabled - perfect for initial setup with contracts like MultiSendCallOnly
• Time-Delayed Security: Schedule addition of new delegates with a configurable delay period for enhanced security
• Instant Removal: Remove delegates immediately without delay when needed
• Comprehensive Coverage: Guards both transaction and module transaction flows in Safe v1.5.0+
• Safe Integration: Seamlessly integrates with Safe smart accounts as a guard contract

Safe Compatibility

Safe Version	Guardrail Contract	Coverage
v1.5.0+	Guardrail.sol	Full (Tx + Module Tx flows)
Demo/Testing	AppGuardrail.sol	Limited (Tx flow only)

About AppGuardrail

• Purpose: Demo and testing only
• Limitations: Only covers normal transaction flow, does not guard module transactions
• Features: Includes helper functions and data structures for frontend integration without requiring indexing services

Project Structure

├── src/                          # Smart contracts
│   ├── Guardrail.sol             # Main guard contract (production)
│   ├── MultiSendCallOnlyv2.sol   # Enhanced MultiSend implementation
│   ├── interfaces/               # Contract interfaces
│   └── test/                     # Test contracts and utilities
├── script/                       # Deployment scripts
├── test/                         # Contract tests
├── guardrail-app/                # React-based Safe App frontend
├── certora/                      # Formal verification specs
└── lib/                          # Dependencies (forge-std, OpenZeppelin, Safe)

Installation

Smart Contracts

# Clone the repository
git clone <repository-url>
cd guardrail

# Install Foundry dependencies
forge install

# Build contracts
forge build

Safe App Frontend

# Navigate to the app directory
cd guardrail-app

# Install dependencies
npm install

# Start development server
npm run dev

Usage

Smart Contract Development

Build Contracts

forge build

Run Tests

forge test

Run Tests with Verbosity

forge test -vv

Format Code

forge fmt

Generate Gas Snapshots

forge snapshot

Deploy Contracts

# Deploy to testnet (example: Sepolia) AppGuardrail
forge script ./script/AppGuardrail.s.sol:AppGuardrailScript --broadcast -vvvv --chain sepolia --rpc-url sepolia --verify

Safe App Development

Start Development Server

cd guardrail-app
npm run dev

The app will be available at http://localhost:3000

Build for Production

npm run build

Run Linting

npm run lint
npm run lint:fix  # Auto-fix issues

Format Code

npm run format
npm run format:check  # Check formatting

Security Considerations

Important

Always review and test thoroughly before using in production environments.

• Time Delays: Configure appropriate delay periods based on your security requirements
• Delegate Verification: Only allow trusted contracts as delegates
• Guard Removal: Guard removal is also time-delayed for security
• Module Guards: Ensure module guard functionality is properly configured for comprehensive protection

Roadmap & Future Improvements

• Module Addition Delays: Implement time delays for module additions to enhance security
• Enhanced MultiSend Integration: Direct decoding of multisend operations to prevent guard bypass attacks
• Gas Optimization: Further optimize contract gas usage