Mindgard

Security testing skills for AI-assisted IDEs and coding agents

25 vulnerability patterns across code execution, prompt injection, data exfiltration, and trust persistence, with black-box and white-box assessment workflows.