Certora Continuous Formal Verification
01Summary
Certora proposes continued security services for Aave protocol with funding of 1,890,000 aUSDC and 9,958 AAVE (totaling ~$2.7M over 12 months). Services include formal verification of new products, security researcher fellowships, community education, and collaboration with auditing firms.
- Type
- treasury allocation
- Funding
- $2,700,000
Key changes
- 12-month stream of 1,890,000 aUSDC to Certora for security services(high)
- 12-month stream of 9,958 AAVE tokens to Certora for security services(high)
- Allocation of six R&D personnel to lead community specification development(medium)
- Free access to Certora Prover SaaS tooling for Aave community members(medium)
- Security researcher fellowships to incentivize top security community members(medium)
- Collaboration partnerships with Sigma Prime, Secureum, Code4rena, and Spearbit for security reviews(medium)
Notable concerns
- Proposal relies on decoded fallback function rather than explicit execute() implementation details
- Past performance shows one critical and two high-severity bugs prevented, but ongoing bug prevention rate unclear
Generated by Kvorum
02Description
Simple Summary
In the last six months, Certora has improved the security of the Aave protocol and played a significant role in increasing the security community’s involvement in Aave’s development. We want to continue to improve Aave’s security and enable agile development of new products. This proposal will give a brief overview of Certora and formal verification, outline what we’ve accomplished in the last six months, and discuss our goals for ensuring Aave’s security.
Abstract
Certora is a security company focused on delivering tools that enable developers to mathematically prove the correctness of DeFi protocols. In the last six months, Certora worked closely with the Aave community, Aave developers, BGD labs, and external communities, including The Secureum 1, to guarantee the security of the Aave protocol and enable massive innovation.
We engaged twenty external security researchers who, along with our team of security engineers and researchers, helped Aave safely develop 16 new products and list 12 new tokens. A total of 51 smart contracts were reviewed which contained over 25,000 lines of Solidity code. One critical and two high-severity bugs were prevented using the hundreds of correctness rules written by the community and the Certora team.
As we continue our collaboration, we plan to improve on the current proposal in five significant ways:
- Allocate more resources
- Improve the Certora Prover
- Develop new open source technology for automatically checking rules written by the community
- Drastically improve the collaboration with auditors and the security research community, and
- Develop a monitoring framework for checking the CVL rules before every transaction.
Motivation
We propose to continue the current engagement. Below are the highlights of what we will accomplish:
- Allocate six R&D personnel to lead the community in specification development for the Aave protocol.
- Write specifications and verify new Aave products and listed tokens.
- Provide free access to our SaaS tooling (Certora Prover) to members of the Aave community.
- Develop and improve our technology to address needs raised by the Aave verification efforts. We will continue to provide new tools to the Aave community as they are developed.
- Develop our education materials and be available to the community to help them learn to use our tools.
- Continue to grow Aave’s security community and incentivize the best of this community to engage with Aave more closely. We propose to award several fellowships to the best security researchers to work closely with Aave and Certora. We have identified leading candidates who contributed to this year’s program.
- Certora will collaborate with Sigma Prime to help them use the Certora Prover as part of their - ongoing security review 1. We will also collaborate with other third parties such as Secureum, who are working to ensure the security of Aave and building the Aave community. Additionally, Code4rena and Spearbit security researchers are proficient in rule writing. For example, Kurt Barry, who works as an independent security researcher through Spearbit, was able to find a 4 year old bug in the MakeDao tool using the Certora Prover. Therefore, for major code revisions, we will conduct Code4rena competitions for finding bugs and writing correctness rules
Specification
A full specification of the payload contract and tests enabling this mandate can be found on the Certora's Github here
The Proposal Payload does the following:
- Creates a 12-month stream of 1,890,000 aUSDC ($1.89M) to the Certora beneficiary address.
- Creates a 12-month stream of 9958 AAVE ($810,000) to the Certora beneficiary address. We've used the 30 days average AAVE price, to calculate the amount of AAVE tokens that corresponds to the $810,000 sum in the proposal. The json file with the price data used is available here
Test Cases
The Proposal Payload has been tested and peer-reviewed by Bored Ghost Developing, including simulations on mainnet of the whole proposal lifecycle.
Link to Test Cases: Here
Implementation
Link to Payload Implementation: Here
Deployed Contracts ProposalPayload = 0x2D2b1Bf70d98ae9A8CC9A3d7a49C2d321eCC6C04
Copyright
Copyright and related rights waived via CC0.
03Mismatch analysis
The proposal description details two specific funding streams (aUSDC and AAVE) to be created via a payload contract, and links to source code, tests, and a deployed payload address that matches the target of the sole decoded on-chain action. However, the decoded calldata only shows a generic execute() call with no arguments — the actual internal logic of the payload (which presumably contains the openStream or similar calls implementing the two streams) is not exposed in the decoded action. Because the calldata gives no visibility into the specific amounts, tokens, or recipients being set, I cannot directly verify the numeric claims (1,890,000 aUSDC, 9958 AAVE) against on-chain parameters. This is a common and expected pattern for Aave governance payloads, where the top-level proposal action is just a delegatecall to a pre-audited, deployed payload contract, and the description, linked GitHub repo, and BGD Labs peer review provide the necessary detail. Given the lack of decoded internal calls, I cannot flag any value or target mismatches with confidence, so no discrepancies are reported. The overall assessment is consistent based on the information available, but confidence is low because the calldata itself does not surface enough detail to fully cross-verify the specific dollar and token amounts described in the proposal text.
Generated by Kvorum
04Actions
- Target
- 0x2d2b…6c04
- Value
- 0 wei
- Function
- fallback()
- Arguments
{}- Calldata
05Tally
Summed from votes
06Voters
No votes recorded yet.
07Forum
No forum thread is confidently linked to this proposal.
08Similar proposals
- aaveexecutedContinuous Formal Verification90%
- aaveexecutedContinuous Security Proposal Aave <> Certora85%
- aaveexecutedAave <> Certora Continuous Security Services85%
- aaveexecutedSecurity Services for Aave Current Infrastructure <> Certora84%
- aaveexecutedSecurity Services for Aave v4 <> Certora84%
- aaveexecutedSigma Prime - Security Assessment Services for Aave73%