MCP Security Toolkit — 10 open-source servers for auditing AI agent environments

[ddjidd564]

Hey everyone! 👋

We just released a suite of open-source MCP servers focused on security auditing, and figured this community might find them useful. There are 10 packages in total—here are a few highlights:

• env-security-scanner — Scans environment variables and config files for credential exposure, secret leaks, and misconfigurations that AI agents might inadvertently expose.
• wallet-security-checker — Audits wallet configurations and detects common deployment security issues in Web3/agentic contexts.
• defi-env-auditor — Checks DeFi agent environments for configuration drift, unsafe defaults, and sandbox escape vectors.

All servers are published on npm and designed to plug straight into any MCP-compatible client (Claude Code, Cursor, etc.). The goal is to help the community build safer agent ecosystems—catch issues early instead of after deployment.

Would love feedback, PRs, or ideas for additional security-focused servers!

Links: npm package listing