Idea
A smart contract auditing platform that combines symbolic and concrete execution with LLMs to reduce false positives and improve defect validation for developers and auditors.
Research Paper
Core Innovation
This paper introduces a pipeline that integrates heuristic detectors, large language models, and formal verification tools to validate Solidity code defects. Unlike prior work that struggles with false positives or misses complex vulnerabilities, this approach uses symbolic and concrete execution to generate proofs and confirm true positives. The chaining of these methods reduces manual verification and improves reliability in smart contract auditing.
Market Size (TAM)
$2–10B TAM for blockchain security and smart contract auditing; $1–2B SAM from enterprises and blockchain platforms. Driven by increasing smart contract adoption and regulatory compliance needs.
Potential Customers & Pain Points
- Blockchain Developers Needing Accurate Vulnerability Detection
- Security Auditors Facing High False Alarm Rates
- Enterprises Requiring Reliable Smart Contract Verification
Business Model
Subscription-based SaaS platform offering automated smart contract auditing with tiered pricing for developers and enterprises.
Competitive Landscape
- MythX
- OpenZeppelin Defender
- CertiK
Implementation Challenges
- High computational cost of symbolic execution
- Inconsistency and cost of large language models
- Integration complexity of multiple tools
Validation Strategy
- Pilot testing with blockchain development teams
- Benchmarking against existing static analysis tools
- User feedback to refine false positive reduction
Research Paper Overview
Validating Solidity Code Defects using Symbolic and Concrete Execution powered by Large Language Models
Summary
This paper presents a novel detection pipeline combining custom Slither-based detectors, Large Language Models, Kontrol, and Forge to reliably detect and prove defects in Solidity smart contracts. It addresses the high false alarm rates of static analysis and LLMs by integrating symbolic and concrete execution methods to validate true positives, reducing manual verification efforts. Experiments show promising results for seven critical defect types, including challenging vulnerabilities like Reentrancy, Complex Fallback, and Faulty Access Control Policies. The approach highlights the benefits and limitations of chaining heuristic analysis with formal verification for automated smart contract auditing.