MCP Security Sandbox

@SirAppSec

Basic information

About MCP Security Sandbox

MCP Security Playground - Hack with MCP Servers, MCP Clients. Try out different vulnerabilities and abuse LLMs and agents in a UI friendly experimentation lab

#security #playground #mcp #agents #llm

MCP Security Sandbox config

No standard config provided

This server doesn't expose a parseable MCP config block in its README. See the repository for install instructions.

Repository

MCP Security Sandbox tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is mcp-security-sandbox?

It is an experimental sandbox and lab for exploring MCP hosts, clients, and servers. It can perform attacks against MCP servers and abuse LLMs.

How to use mcp-security-sandbox?

Install dependencies using uv install, create a virtual environment with uv venv, activate it, then run the server with uv run -- src/mcp-security-sandbox/mcp/github/server.py and launch the Streamlit frontend with streamlit run src/mcp-security-sandbox/frontend/MCP_Chat.py. Ensure Ollama is installed and its URL is set in the client initializations.

Key features of mcp-security-sandbox?

Experimental sandbox for MCP exploration and attacks
Integrates a GitHub retrieval MCP server into a chat agent
Supports chaining with Burp Suite MCP Server
Provides a Streamlit-based chat interface
Allows abuse of LLMs for attack scenarios
Uses Ollama for local LLM integration

Use cases of mcp-security-sandbox?

Test security of MCP servers against attacks
Simulate malicious MCP server behavior
Experiment with chaining multiple MCP servers
Educate on MCP security vulnerabilities

Frequently asked questions

What dependencies are required?

Ollama must be installed and its URL configured in the client initializations. The project uses `uv` for package and environment management.

How do I start the frontend?

Run `uv install`, `uv venv`, `source .venv/bin/activate`, then `uv run -- src/mcp-security-sandbox/mcp/github/server.py` and `streamlit run src/mcp-security-sandbox/frontend/MCP_Chat.py`.

Does the sandbox support dynamic loading of MCP servers?

No, dynamic loading is on the roadmap but not yet implemented; currently servers are statically defined.

What is the purpose of this project?

It is an experimental sandbox for exploring MCP hosts, clients, and servers, and for performing attacks against MCP servers and abusing LLMs.

Can I integrate Burp Suite?

Yes, the README shows using Burp Suite MCP Server to

Comments

More Developer Tools MCP servers

MeroFoundry

MERO Consulting Inc

MeroFoundry is a remote MCP server available at https://platform.merofoundry.com/mcp.

Launchdog

Devin Starks

Launch your product from your AI agent. Connect Claude Code, Codex, Cursor or ChatGPT to Launchdog over MCP and your agent drafts the listing, picks a launch week and books it. 50 launches a week, ranked only by human vo

Agent Conductor

icohangar-ops

Agent conductor — AGENTS.md contract parser, SKILL.md registry, CHP decision engine over MCP

Modelglass

Modelglass turns "which model should I use?" into a lookup instead of a guess. Seven tools give your agent live pricing across image, language, video, and audio models. Sourced and versioned, not scraped once and forgott

Yocoolab

MCP server for AI Agents — exposes Yocoolab feedback threads and design selections as tools