A coding-agent skill that runs multi-phase security audits with isolated agents and independently verified, machine-readable findings.
Agent harness system of skills, agents, hooks, rules, memory, and security scanning for Claude Code, Codex, Cursor, and other AI coding tools.