Add mastyf.ai — MCP runtime security proxy to community projects

[mastyf-ai]

Adds mastyf.ai to the community projects list.

mastyf.ai is an open-source (MIT) runtime security proxy for the Model Context Protocol. It sits transparently between MCP clients and servers, intercepting every tools/call and enforcing security policy before execution.

• 18 attack classes covered: prompt injection, SSRF, shell injection, SQL injection, credential exfiltration, polyglot attacks, deserialization, path traversal, and more
• YAML policy engine: configurable rules with allow/deny patterns
• Trust scoring: scores npm MCP packages with 0–100 trust badges
• Defense Fabric: 6-phase pipeline (lifecycle → pre-guard → hooks → policy → semantic → spend)
• Active development: publishes to PyPI, Docker (ghcr.io), npm SDKs, and a cloud dashboard at mastyf-live.vercel.app