Automates multi-stage Solidity smart contract audits for CI/CD and evaluation pipelines using Claude Code.
This skill provides a specialized, non-interactive variant of the Solidity function audit pipeline, optimized for automated testing and evaluation via the `claude -p` mode. It streamlines the auditing process by removing all interactive pause points, reading design decisions from a local GROUND_TRUTH.md file, and executing a rigorous 7-stage analysis. From state variable mapping to cross-cutting adversarial sequence analysis, this skill is designed for developers who need to integrate consistent, AI-driven security assessments into their headless automated workflows or performance evaluation suites.
主要功能
01Background execution of parallel analysis agents for state, access control, and external calls.
02Automated discovery of contract domains, functions, and DeFi patterns.
03Non-interactive 7-stage automated audit pipeline for headless execution.
04Comprehensive synthesis of findings with automated severity tallies and executive summaries.
050 GitHub stars
06Pre-seeds design decisions from ground-truth files to eliminate manual configuration.
使用场景
01Benchmarking AI performance on large-scale Solidity codebases without manual intervention.
02Integrating automated smart contract security audits into CI/CD pipelines.
03Generating standardized security reports for multi-contract DeFi protocols automatically.
