2026-11-05 –, Lotus Stage 2
We show how AI agents can turn informal requirements into formally verified Aave V4 spokes, using the Certora Prover as a correctness guardrail. Attendees will learn how “verified vibe coding” enables rapid development without trusting LLM-generated code to be correct.
Aave V4 spokes enable easy deployment of various lending markets, each with different collateral management, risk premiums, liquidations, per reserve risk configuration, and oracle integration, without having to manually manage liquidity. We show how to leverage agents to develop new secure spokes. The core idea is that the user provides an informal definition of correctness for these applications. Our agentic pipeline translates these informal definitions into a formal specification of the application’s safety properties/invariants. Once we have these formal specifications, we can Vibe-code the spoke implementation guardrail with the Certora Prover. The LLM cannot introduce bugs into our generated implementation because the Certora Prover will reject any potential implementation that does not satisfy the formal specification.
Mooly Sagiv is a professor and chair of Computer Sciences at Tel-Aviv University and a CEO and co-founder of Certora. He is a leading researcher in large-scale (inter-procedural) program analysis and one of the key contributors to shape analysis. His fields of interest include programming languages, compilers, abstract interpretation, profiling, pointer analysis, shape analysis, interprocedural dataflow analysis, program slicing, and language-based programming.