Shieldon Auditor

This tool audits a single input codeSnippet string containing source code or text for security risks, including hard-coded credentials, exfiltration endpoints, dangerous commands or APIs, obfuscation patterns, and suspicious control flow or side effects. It semantically evaluates the snippet and produces a structured JSON security assessment with findings and a normalized risk_score from 0 (critical) to 100 (safe). The tool accepts only the codeSnippet payload as input, and its output must strictly validate against the system-provided structured output schema used by the caller.