agent-security-scanner-mcp

Security layer for AI agents: blocks prompt injection, detects fake packages, scans vulnerabilities.