Corporate Multisig: Integrating Safe{Wallet} for DeFi Projects
Most DeFi teams overlook the nuances of setting up a multisig wallet. Safe{Wallet} (formerly Gnosis Safe) is the de facto standard a widely-used standard in the EVM ecosystem. Over $100B in assets are under its management, the code is heavily audited, and the built-in UI at the official interface gets you started in minutes. But a proper multisig setup is not just clicking "create": you need to choose the right owners, threshold, consider gas costs, and optionally attach a Guard for policy enforcement. Our team has 6+ years of experience with Safe multisig wallet deployments, managing over 50 configurations for DeFi projects with total treasury value exceeding $500M. We provide a guaranteed secure setup with a 30-day warranty and ongoing support. Compared to legacy multisig solutions, Safe provides 10x more security with its module ecosystem and Guard policy enforcement. Additionally, our standard setup costs start from $2,000, and by deploying on L2 you can save up to 90% on gas compared to Ethereum mainnet. We've helped configure treasuries for 30+ DeFi projects – and we know the pitfalls that await inexperienced users.
How to Choose Threshold and Owners?
Owners are addresses whose signatures the contract accepts. Each owner can be an EOA or a smart contract (e.g., another Safe). Use hardware wallet addresses for owners, not browser keys from MetaMask.
Threshold (M-of-N) is the number of signatures required to execute a transaction. For a team treasury: 3-of-5 or 2-of-3. Do not use 1-of-N for real funds – it's a single point of failure. Do not use N-of-N – losing one key locks everything. Comparison:
| Threshold | Security | Operability | Suitable for |
|---|---|---|---|
| 1-of-2 | low | high | personal wallets, test |
| 2-of-3 | medium | good | startups, small teams |
| 3-of-5 | high | acceptable | medium DAOs, funds |
| 4-of-7 | very high | low | large treasuries |
Chain – Safe deploys separately on each network. The same Safe address can be obtained on different networks via the Safe Factory with the same saltNonce – useful for a multichain treasury with a single address.
Why Can a Multisig Be Vulnerable?
Without proper configuration, a multisig can become vulnerable: a wrong threshold leads to locked funds, missing a Guard allows unauthorized transfers, and weak owners lead to key theft. Our experience includes setting up Safes that have been running for 5+ years in DeFi projects, auditing configurations, and formally verifying Guard policies. A common mistake is storing all private keys on one Ledger device, only switching accounts. This violates the multisig principle: if the device is compromised, the attacker gets all keys. Each owner must be on a separate physical device. Lacking a Safe Guard can lead to losses exceeding $100k in a targeted attack.
What Risks Does Safe Guard Mitigate?
Safe Guard is a smart contract that is called before every transaction from the Safe. It allows adding policies: whitelist of recipient addresses, amount limits, prohibition of certain functions. Without a Guard, any 2-of-3 owners can send any transaction anywhere. With a Guard, an additional check logic is enforced that cannot be bypassed even with full quorum. The implementation is standardized in ISafeGuard. Describe your project, and we'll help choose the optimal Safe configuration for your team.
Signing and Executing Transactions
A transaction is created in the Safe UI or via SDK. Each owner signs off-chain (no gas). After collecting threshold signatures, anyone can call execTransaction on-chain and pay gas. This does not have to be an owner.
| Operation | Gas cost (Ethereum mainnet) |
|---|---|
| Deployment Safe | ~280,000 gas |
| execTransaction (2-of-3) | ~120,000–150,000 gas |
| Adding owner | ~80,000 gas |
| Changing threshold | ~50,000 gas |
On L2 (Arbitrum, Base, Optimism), gas is orders of magnitude lower – Safe is significantly cheaper there. Gas savings when moving to L2 can reach 90%, which for an active treasury with 50+ transactions per month means tens of thousands of dollars in annual savings. Determine the optimal configuration for your budget – book a technical discussion.
Setup Process
- Analytics: team interviews, role definition, signing scenarios.
- Design: choose owners, threshold, need for a Guard. Create configuration scheme.
- Deployment: create Safe via SDK or UI on the required networks with a deterministic address.
- Guard Setup: develop and deploy a custom Guard with checks (whitelist, limits).
- Test: simulate transactions in Tenderly, verify all policies.
- Go-Live: confirm operability, hand over documentation.
- Support: two weeks post-deployment, help with first transactions.
Programmatic creation via Safe{Core} SDK:
import { SafeFactory } from '@safe-global/protocol-kit' const safeFactory = await SafeFactory.create({ ethAdapter }) const safe = await safeFactory.deploySafe({ safeAccountConfig: { owners: ['0xAlice...', '0xBob...', '0xCarol...'], threshold: 2, }, saltNonce: '0x123' // for deterministic address }) Security Recommendations
Each owner stores the key on a separate hardware wallet. Backup seed phrases physically, in different locations. When a team member changes, add a new owner and remove the old one via a Safe transaction (requires current quorum). Never set threshold equal to total owners: losing one key means losing access.
What Is Included
Deliverables:
- Designing owners/threshold scheme for your team (2–20 participants)
- Deploying Safe on required networks (Ethereum, L2, sidechains) with a deterministic address
- Configuring Safe Guard with custom policies (whitelist, limits, function restrictions)
- Integration via Safe{Core} SDK for programmatic management
- Configuration documentation, team training, access handover
- Support for two weeks after handover
Setting up Safe for a team of 2 to 20 participants with correct policies takes 1–3 days. Contact our engineers for a consultation on multisig setup.