sammyboi81

MCP Agent Security — Security skill for Claude Code

Security community

Securing MCP servers, tool use, and autonomous coding agents - tool poisoning, confused deputy, token passthrough, sandboxing.

How to install MCP Agent Security

Installs to ~/.claude/skills/sammyboi81-bulwark-mcp-mcp-agent-security/SKILL.md

Terminal

mkdir -p ~/.claude/skills/sammyboi81-bulwark-mcp-mcp-agent-security && curl -fsSL https://raw.githubusercontent.com/sammyboi81/bulwark-mcp/HEAD/skills/mcp-agent-security.md -o ~/.claude/skills/sammyboi81-bulwark-mcp-mcp-agent-security/SKILL.md

Restart Claude Code, or start a new session, for it to be picked up.

What MCP Agent Security does

name: mcp-agent-security description: Securing MCP servers, tool use, and autonomous coding agents - tool poisoning, confused deputy, token passthrough, sandboxing. tags: ai, mcp, agents

Using MCP servers and agent tools

Every MCP server is code running with your privileges plus text that steers your agent. Vet it like a dependency: known maintainer, source reviewed, pinned version/commit, minimal permissions, no auto-update (rug-pull risk).
Tool poisoning: tool names, description

Alternatives in Security

Trail Of Bits Claude-code-config — Opinionated production defaults from a top security firm: sandboxing, permissions, hooks, skills, MCP server c 1.6k ★
Google MCP Security Servers — Security Operations and Threat Intelligence MCP servers 453 ★
Agentseal — Security toolkit for AI agents 344 ★