Bắt đầu làm việc tại công ty sử dụng MCP quy mô lớn. Đang xây dựng mô hình mối đe dọa. Bạn đã biết về vấn đề tiêm nhiễm gián tiếp và phép dùng công cụ trái phép, nhưng bạn tìm kiếm những "bẫy đẫy" nào? Những vấn đề bảo mật nào thực sự đang gây khó khăn cho bạn khi triển khai MCP trong doanh nghiệp? #BảoMậtMCP #AnToànAI #threatmodeling #LocalLLaMA #MCPSecurity #AISecurity #MôHìnhĐeDọa

https://www.reddit.com/r/LocalLLaMA/comments/1py3uru/securing_mcp_in_production/

We deployed MCP honeypots to understand how threat actors engage with AI middleware exposed to the internet. What we observed was unexpected. Full analysis: https://www.greynoise.io/blog/deploying-mcp-honeypots

#GreyNoise #AI #AISecurity #MCP #MCPSecurity #Cybersecurity #ThreatIntel

What GreyNoise Learned from Deploying MCP Honeypots

GreyNoise deployed MCP honeypots to see what happens when AI middleware meets the open internet — revealing how attackers interact with this new layer of AI infrastructure.

Good read for security engineers trying to wrangle MCP tooling.

#mcp #mcpsecurity #aisecurity

https://semgrep.dev/blog/2025/a-security-engineers-guide-to-mcp/

A Security Engineer's Guide to MCP

MCP is quickly becoming the API standard for AI coding agents. That means new attack surfaces — and security engineers need to know how to test them safely.

Semgrep