[RFC] Safe Flash Loan Module

Safe Development Modules
1

Hey everyone!

We’d like to gather feedback on a potential Safe Module to enable flash loans using funds held in Safe wallets.

Problem statement

Today, flash loan liquidity is fragmented across multiple DeFi protocols (Aave, Uniswap, Balancer, etc.), each with different interfaces, fee structures, and supported tokens. Integrators — arbitrageurs, liquidators, and other searchers — need to integrate with multiple protocols and manually check where liquidity is available for each token they need. There’s no unified way to discover flash loan liquidity across the ecosystem.

Beyond fragmentation, many tokens simply are more liquid on Safe wallets than on these protocols to support large flash loan amounts. Take the COW token as an example: on mainnet, there’s roughly 400k USD available for flashloans on Balancer and around 1M USD on Uniswap — but the CoW DAO treasury holds over 40M USD. In Base, the gap is even stronger: ~30k USD on Uniswap compared to 175k USD sitting in a single Safe wallet. This pattern repeats across many governance and protocol tokens where the majority of supply lives in multisigs, not in DeFi liquidity pools.

Flash loans have become essential for DeFi infrastructure. But this is fundamentally a race-to-the-bottom market: integrators optimize for just two factors — fees and liquidity. Aave charges 0.09% and doesn’t support many tokens, making it less attractive for most use cases. Morpho is being more used because of its larger variety of tokens (source). This dynamic means that any new liquidity source with competitive fees can capture significant volume — especially for long-tail assets where existing options are limited.

This means a lot of capital sitting in DAO treasuries is effectively idle, even though it could be generating yield while remaining fully accessible. What if Safe owners could make their tokens available for flash loans in exchange for a fee, without sacrificing custody or security guarantees?

We’re aware this hasn’t been a frequently requested feature from Safe users so far. This RFC is part of an early market research & validation step to understand whether treasuries and integrators actually want this before anyone invests heavily in building it.

The idea

We’re exploring a design with three components:

Safe Module

A module to handle the flash loan logic, following the standard ERC-3156 specification for easy integration with existing tooling. It enforces atomic execution — if the funds + fee are not returned in the same transaction, the state reverts. By following an established standard, any existing flash loan integration can work with Safe-based liquidity with minimal changes.

Safe App UI

A user-friendly interface for Safe owners to enable/disable flash loans for specific tokens, set fee percentages, and monitor stats (fees earned, usage history, etc.).

Liquidity Discovery Indexer

An indexer that aggregates available flash loan liquidity across all participating Safes. This addresses the fragmentation problem head-on: with a simple API call, integrators could discover which addresses offer flash loans for a given token, compare fees, and route to the best available liquidity. The goal is to make Safe-based flash loans as easier to integrate than existing protocols even with liquidity within multiple wallets.

This module isn’t meant to compete with Aave on ETH/USDC liquidity; rather, it can help serve long-tail assets and potentially lower fees across the ecosystem, making it more efficient overall.

At this stage, we’re mainly trying to understand whether this is something the Safe community would want and how it should be shaped.

Key points

  • Long-tail assets not served by existing flash-loan markets

  • Any revenue is additive to whatever the treasury is already doing;

  • Part of this RFC is to validate whether integrators would actually use this liquidity

Questions for the community

As a flash loan user:

  • Would you realistically use Safe-based flash loans for long-tail assets?

  • Would a discovery API that aggregates Safe-based flashloan liquidity be useful for your workflows?

As a DAO:

  • What help would you need to set the fee percentage?

  • Do you think that caps (max amount to loan) is an important feature?

  • Are there specific attack vectors you’d be particularly worried about (beside flashloan not being paid)?

General:

  • Is this something that would be valuable for the Safe ecosystem in general, or does it feel out of scope?

  • Are we missing prior existing work we should build on instead?

About us

We’re web3 builders partnering with companies and DAOs to make blockchain interactions simple and user-friendly. Our mission is to bridge the gap between blockchain’s potential and current user experiences, helping onboard the next billion users to blockchain.

We’ve previously:

We’d love to hear your thoughts, questions, and concerns before we dive deeper into design and implementation.

1 Like