Аудит безопасности CLI: инъекции и экранирование

security

security-guardian is a Claude Code skill that performs comprehensive security analysis of the RTK CLI tool, targeting command injection, shell escaping, hook security, and malicious input handling. It triggers automatically after filter changes, shell command execution logic updates, or hook modifications, and supports manual invocation before releases or after security-sensitive code changes. The skill implements RTK's full threat model: command injection and shell escaping rated CRITICAL (RCE risk), hook injection rated HIGH, malicious output handling rated MEDIUM, and path traversal rated LOW. It provides Rust and Bash code audit patterns, security test suites for injection defense, and guidance on safely parsing untrusted output from external commands like git, cargo, and gh. Built for developers securing CLI proxies that execute shell commands based on user-controlled input.