Google Cloud releases preview of "CodeMender," an AI that autonomously detects code vulnerabilities, validates risks in a sandbox, and automates fixes

Google Cloud has previewed CodeMender, an AI agent that automates vulnerability detection, sandbox validation, and fix generation.

The tool identifies security issues in code, checks exploitability in a sandbox, and then proposes repair code automatically.

By combining static analysis, code review, and DevSecOps workflows, it aims to reduce false positives and speed up prioritization.

It supports Visual Studio Code, Antigravity, CI/CD environments, and major languages including C/C++, Go, Java, Python, Ruby, Rust, and TypeScript.