petrovicistefan/mcp-security-guard

Overview

Security scanning and guardrails for MCP servers and AI agents. Built in TypeScript, this open-source tool aims to help developers identify and mitigate security issues in Model Context Protocol integrations.

Best for

Best for
Developers building MCP servers or AI agents who need security scanning.

Use cases

• Scan MCP servers for security vulnerabilities
• Enforce guardrails on AI agent actions
• Integrate security checks into MCP development workflows

How to use

Tools exposed

• list_mcp_servers
• audit_mcp_config
• audit_server_tools
• pin_tools
• export_bom
• audit_agent_context
• pin_context
• analyze_tool_definitions
• check_supply_chain
• apply_fixes
• security_dashboard
• generate_policy
• query_audit_log
• adversarial_test

Tested with

Claude Desktop, Claude Code, Cursor, Windsurf, Cline, VS Code

Example client config

{ "mcpServers": { "mcp-security-guard": { "command": "npx", "args": ["-y", "mcp-security-guard"] } } }

Notes

Security scanning and guardrails for MCP servers and AI agents. Built in TypeScript, this open-source tool aims to help developers identify and mitigate security issues in Model Context Protocol integrations.

0 stars on GitHub. Last updated 2026-10-09. Licensed MIT.

Use cases

• Scan MCP servers for security vulnerabilities
• Enforce guardrails on AI agent actions
• Integrate security checks into MCP development workflows

Pros

• Focused specifically on MCP security
• TypeScript codebase for easy integration
• Open source with public repository

Cons

• No GitHub stars yet, indicating early adoption
• Limited documentation or community support
• Unclear feature set beyond the description

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.