SecureVibes Introduces Multi-Language Vulnerability Scanner Powered by Claude AI

SecureVibes has unveiled a multi-language vulnerability scanner powered by Claude AI, designed to deliver intelligent security analysis across eleven programming languages. The platform uses Claude’s multi-agent architecture with five specialised AI agents—four core agents and one optional dynamic testing agent—to provide context-aware analysis with concrete evidence such as file paths and line numbers.

It automatically detects vulnerabilities in Python, JavaScript, TypeScript, Go, Ruby, Java, PHP, C#, Rust, Kotlin, and Swift, and handles polyglot projects by combining exclusion rules across languages, including venv, node_modules and vendor directories. The detection process unfolds in distinct phases, beginning with an Assessment Agent, then a Threat Modeling Agent applying STRIDE, followed by a Code Review Agent and a Report Generator, with an optional DAST Agent for dynamic validation.

Users can run the full scan or per-agent, with a CLI that supports Markdown, JSON and terminal tables, and features per-agent model selection and streaming mode for real-time progress. Version 0.3.1 is available on PyPI, with the platform transmitting source code and relative file paths to Claude API while excluding absolute paths, environment variables and git metadata.