mcp-security-hub

Why we included this project

Security teams that already run Claude or another MCP-capable assistant will get the most out of this project, because it lets them drive offensive tooling through natural language. It packages 38 Dockerized MCP servers that wrap familiar utilities like Nmap, Nuclei, SQLMap, Ghidra, radare2, and Hashcat, so a pentester can ask for a port scan, a template-based vulnerability check, or a binary analysis without juggling separate command-line invocations. Coverage spans reconnaissance, web security, binary analysis, cloud auditing, secrets detection, and blockchain security, which suits both broad assessments and focused engagements. The containers run as non-root users and are scanned with Trivy, and Docker Compose orchestration plus ready-made Claude configuration templates keep setup to building the images and pointing Claude at them. If you are weighing how to bring AI into your security workflow, this is a good place to start.