Web3 Security Triager (AuditAgent & AgentArena)

Skills

About the Role

You will validate AI-generated vulnerability reports, triage and classify findings, and filter out false positives to ensure customers receive actionable security insights. You will run AuditAgent in live audit competitions, write Proof of Concept exploit code for valid findings, submit and track results, and provide feedback to improve detection quality. You will document findings and help build benchmarks, propose triage strategies, and suggest automation to raise overall detection accuracy.

Requirements

• Solid understanding of Web3 security and common smart contract vulnerability classes such as reentrancy, access control, and oracle manipulation
• Proficiency in Solidity with ability to read and reason about contract logic
• Proficiency with AI coding tools (e.g., Cursor, Claude Code) to accelerate workflows
• Ability to write Proof of Concept exploits demonstrating valid bugs
• Strong attention to detail and clear severity reasoning
• Proactive and creative mindset
• Nice to have: experience with Solana and Rust smart contract security
• Nice to have: prior participation in audit competitions (Code4rena, Sherlock, Immunefi)
• Nice to have: familiarity with security tools such as Slither and Foundry

Responsibilities

• Triage AI-generated findings and validate vulnerability reports
• Filter false positives and classify severity with clear reasoning
• Run AuditAgent on live audit competitions and validate outputs
• Write Proof of Concept exploit code for valid findings
• Submit validated findings and track contest results
• Provide feedback to product and engineering to improve detection quality
• Document results and produce internal reports and public content

Benefits

• Remote-first, globally distributed team