CTF MCP

CTF-MCP

[English](#english) | [中文](#中文)

---

English

A Model Context Protocol (MCP) server providing comprehensive tools for CTF (Capture The Flag) challenges.

⚠️ Security Warning

IMPORTANT: This tool is designed for authorized security testing only.

• ✅ Authorized Use: CTF competitions, penetration testing with written permission, security research, educational purposes
• ❌ Prohibited Use: Unauthorized access, malicious attacks, illegal activities

Before using this tool:

1. Ensure you have proper authorization
2. Read [SECURITY.md](SECURITY.md) for detailed security guidelines
3. Review [BESTPRACTICES.md](BESTPRACTICES.md) for usage recommendations
4. Understand the risk levels of tools you plan to use

Risk Levels:

• 🟢 LOW: Information gathering (safe)
• 🟡 MEDIUM: Vulnerability detection (requires authorization)
• 🟠 HIGH: Exploitation tools (explicit authorization required)
• 🔴 CRITICAL: RCE/Deserialization payloads (written authorization required)

Security Audit Results:

• 🔒 Total Security Decorators: 33 dangerous operations protected
• 🔴 CRITICAL: 14 tools (RCE, deserialization)
• 🟠 HIGH: 13 tools (exploitation, injection)
• 🟡 MEDIUM: 6 tools (weak crypto, code generation)
• See [SECURITY.md](SECURITY.md) for complete tool lists and descriptions

By using CTF-MCP, you agree to use it responsibly and legally.

Features

Crypto (53 tools)

• Base encoding (Base64, Base32, Base58, Base85)
• Classical ciphers (Caesar, Vigenere, Atbash, Affine, Rail Fence, Playfair, Hill)
• XOR operations and analysis
• Modern crypto (AES, DES, RC4)
• Hash functions and cracking
• RSA attacks (Wiener, Hastad, Common Modulus, Franklin-Reiter)
• Frequency analysis

Web (46 tools)

• SQL Injection (Union, Error, Blind, Time-based)
• XSS payloads and filter bypass
• SSTI (Server-Side Template Injection)
• SSRF, XXE, Command Injection
• JWT attacks (None algorithm, Key confusion)
• Deserialization (PHP, Python Pickle, Java, Node.js)
• HTTP Smuggling, GraphQL, WebSocket
• OAuth, CORS, Cache Poisoning

Pwn (27 tools)

• Shellcode generation (x64/x86)
• Cyclic pattern creation and offset finding
• ROP gadgets and chain building
• Format string exploits
• Heap exploitation (Tcache, Fastbin, House of Force/Spirit)
• Stack pivot techniques
• SROP (Sigreturn-oriented programming)
• Libc database and one_gadget

Reverse

• Disassembly helpers
• String extraction
• Binary analysis

Forensics

• File carving
• Memory analysis
• Network forensics

Misc

• Encoding/decoding utilities
• Steganography helpers

Installation

pip install -e .

Usage

As MCP Server

ctf-mcp

In Python

from ctf_mcp.tools.crypto import CryptoTools
from ctf_mcp.tools.web import WebTools
from ctf_mcp.tools.pwn import PwnTools

crypto = CryptoTools()
web = WebTools()
pwn = PwnTools()

# Generate SQL injection payloads
print(web.sql_payloads("mysql", "union"))

# Base64 encode
print(crypto.base64_encode("Hello CTF"))

# Create cyclic pattern
print(pwn.pattern_create(100))

MCP Configuration

Add to your MCP client configuration:

{
  "mcpServers": {
    "ctf-mcp": {
      "command": "ctf-mcp"
    }
  }
}

Requirements

• Python 3.10+
• mcp >= 1.0.0

---

中文

一个为 CTF（夺旗赛）挑战提供全面工具的 MCP（模型上下文协议）服务器。

⚠️ 安全警告

重要提示：本工具仅用于授权的安全测试。

• ✅ 授权使用：CTF 竞赛、获得书面许可的渗透测试、安全研究、教育目的
• ❌ 禁止使用：未经授权的访问、恶意攻击、非法活动

使用本工具前：

1. 确保您已获得适当的授权
2. 阅读 [SECURITY.md](SECURITY.md) 了解详细的安全指南
3. 查看 [BESTPRACTICES.md](BESTPRACTICES.md) 了解使用建议
4. 了解您计划使用的工具的风险级别

风险级别：

• 🟢 低风险：信息收集（安全）
• 🟡 中风险：漏洞检测（需要授权）
• 🟠 高风险：漏洞利用工具（需要明确授权）
• 🔴 严重风险：RCE/反序列化 payload（需要书面授权）

安全审计结果：

• 🔒 安全装饰器总数: 33个危险操作受保护
• 🔴 高危: 14个工具（远程代码执行、反序列化）
• 🟠 高风险: 13个工具（漏洞利用、注入）
• 🟡 中风险: 6个工具（弱加密、代码生成）
• 查看 [SECURITY.md](SECURITY.md) 获取完整工具列表和说明

使用 CTF-MCP 即表示您同意负责任且合法地使用它。

功能特性

密码学工具 (53个)

• Base 编码 (Base64, Base32, Base58, Base85)
• 古典密码 (凯撒、维吉尼亚、Atbash、仿射、栅栏、Playfair、Hill)
• XOR 运算与分析
• 现代加密 (AES, DES, RC4)
• 哈希函数与破解
• RSA 攻击 (Wiener、Hastad、共模攻击、Franklin-Reiter)
• 频率分析

Web 安全工具 (46个)

• SQL 注入 (联合查询、报错注入、布尔盲注、时间盲注)
• XSS 载荷与过滤器绕过
• SSTI (服务端模板注入)
• SSRF、XXE、命令注入
• JWT 攻击 (None 算法、密钥混淆)
• 反序列化 (PHP, Python Pickle, Java, Node.js)
• HTTP 走私、GraphQL、WebSocket
• OAuth、CORS、缓存投毒

二进制利用工具 (27个)

• Shellcode 生成 (x64/x86)
• 循环模式创建与偏移查找
• ROP gadgets 与链构建
• 格式化字符串漏洞利用
• 堆利用 (Tcache, Fastbin, House of Force/Spirit)
• 栈迁移技术
• SROP (Sigreturn 导向编程)
• Libc 数据库与 one_gadget

逆向工程

• 反汇编辅助
• 字符串提取
• 二进制分析

取证分析

• 文件雕复
• 内存分析
• 网络取证

杂项工具

• 编码/解码工具
• 隐写术辅助

安装

pip install -e .

使用方法

作为 MCP 服务器

ctf-mcp

在 Python 中使用

from ctf_mcp.tools.crypto import CryptoTools
from ctf_mcp.tools.web import WebTools
from ctf_mcp.tools.pwn import PwnTools

crypto = CryptoTools()
web = WebTools()
pwn = PwnTools()

# 生成 SQL 注入载荷
print(web.sql_payloads("mysql", "union"))

# Base64 编码
print(crypto.base64_encode("Hello CTF"))

# 创建循环模式
print(pwn.pattern_create(100))

MCP 配置

添加到你的 MCP 客户端配置：

{
  "mcpServers": {
    "ctf-mcp": {
      "command": "ctf-mcp"
    }
  }
}

环境要求

• Python 3.10+
• mcp >= 1.0.0

---

Author / 作者

Coff0xc

License / 许可证

MIT License - see [LICENSE](LICENSE)

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: Coff0xc
• Source: Coff0xc/CTF-MCP
• License: MIT

Install and usage instructions live in the source repository linked above.