Thomas Pani

Thomas Pani

@thpani.bsky.social

Protocol Security Researcher :: https://blltprf.xyz | prev: @informalinc, @tuwien.at fuzzing, formal verification, bespoke security engineering

17 followers13 following9 posts

Getting ready for my #Solidity #fuzzing workshop at @protocol.berlin 🔥 Join me to build a practical EVM smart contract fuzzer in Python and see how a Formal Methods viewpoint can guide your approach to smart contract security testing. #EVM #SmartContracts #Security

1/ Distributed algorithms like blockchain consensus protocols are notoriously difficult 🤯 Too many combinations, too easy to miss bugs—fault-tolerance makes it worse. Tools like TLA+ & Apalache, Alloy, and CVC5 can help! Read our report: [arxiv.org/abs/2501.07958] 🧵

arxiv.orgTechnical Report: Exploring Automatic Model-Checking of the...We investigate automated model-checking of the Ethereum specification, focusing on the Accountable Safety property of the 3SF consensus protocol. We select 3SF due to its relevance and the unique...
View full thread

5/ Insights (3): Ultimately, checking the Alloy model with @arminbiere.bsky.social's Kissat SAT solver enabled verification of the largest state-space. Our approach demonstrates the power of combining tools to push the limits of what’s achievable in distributed protocol verification. ⚡️

1/ Distributed algorithms like blockchain consensus protocols are notoriously difficult 🤯 Too many combinations, too easy to miss bugs—fault-tolerance makes it worse. Tools like TLA+ & Apalache, Alloy, and CVC5 can help! Read our report: [arxiv.org/abs/2501.07958] 🧵

arxiv.orgTechnical Report: Exploring Automatic Model-Checking of the...We investigate automated model-checking of the Ethereum specification, focusing on the Accountable Safety property of the 3SF consensus protocol. We select 3SF due to its relevance and the unique...
View full thread

3/ Insights (1): Specs in TLA+, Alloy, and SMT at various abstraction levels let us exhaustively check Accountable Safety for parameters reasonably large for verification. Extensive experiments, lasting up to 16 days, found no counterexamples—boosting confidence in the protocol’s safety. 🚀

4/ Insights (2): Combining tools was pivotal. Refining the spec in TLA+ deepened our understanding of the protocol and shaped abstractions. Generating configurations like justified/finalized checkpoints was much easier with TLA+ and Alloy than with the initial Python spec. 🧠

1/ Distributed algorithms like blockchain consensus protocols are notoriously difficult 🤯 Too many combinations, too easy to miss bugs—fault-tolerance makes it worse. Tools like TLA+ & Apalache, Alloy, and CVC5 can help! Read our report: [arxiv.org/abs/2501.07958] 🧵

arxiv.orgTechnical Report: Exploring Automatic Model-Checking of the...We investigate automated model-checking of the Ethereum specification, focusing on the Accountable Safety property of the 3SF consensus protocol. We select 3SF due to its relevance and the unique...

You are up to date.