AdaptixC2 MCP Server
Enables LLM agents to automate red teaming operations with the AdaptixC2 framework via MCP protocol, supporting post-exploitation tasks and BOF integration.
README
AdaptixC2 MCP Server
✨ Overview
AdaptixC2 MCP Server enables LLM agents to interact with and automate operations with the AdaptixC2 framework via the MCP protocol.
It allows AI-driven red teaming, automation of post-exploitation tasks, and controlled tool execution within a C2 environment.

🚀 Installation and Setup
Python 3.10+ and the uv package manager (curl -LsSf https://astral.sh/uv/install.sh | sh) are required.
- Clone the repository:
git clone https://github.com/Faceless0x7/AdaptixC2-MCP-Server
cd AdaptixC2-MCP-Server
- Create an isolated virtual environment and install dependencies:
uv venv
source .venv/bin/activate
uv pip install -r requirements.txt
- Configure connection parameters: Copy the configuration template and edit it with your details (IP, port, C2 credentials, endpoint):
cp .env.example .env
🤖 Connecting to LLMs
Add the server startup parameters to your MCP client configuration.
Example: Gemini CLI
Create the configuration file:
mkdir -p .gemini
nano .gemini/settings.json
Example configuration for Gemini CLI:
{
"mcpServers": {
"adaptixc2": {
"command": "uv",
"args": [
"--directory",
"/home/kali/Desktop/AdaptixC2-MCP-Server",
"run",
"server.py"
]
}
}
}
🛡️ BOF Integration (Optional)
The server supports BOFs. To activate this functionality, additional setup steps are required on both the Teamserver and the MCP side.
1. Build and Install Extension-Kit
Clone and compile the modules from the official Extension-Kit repository:
git clone https://github.com/Adaptix-Framework/Extension-Kit
cd Extension-Kit
make
2. Configure profile.yaml in AdaptixC2
In the profile.yaml configuration file of your AdaptixC2 Teamserver, you must specify the path to the compiled file from the first step (it is important to use an absolute path):
axscripts:
- "/full/path/to/Extension-Kit/extension-kit.axs"
For more details about the main C2 server: AdaptixC2 Documentation
3. Restricting AI Access via bofs.yaml
To minimize risk and avoid excessive tool exposure, the server enforces a strict default-deny model for BOF modules.
Edit the bofs.yaml file in the root directory of AdaptixC2 MCP Server. You must explicitly list which commands the AI is allowed to leverage:
# Example: Granting access only to two required BOFs in AD
AD-BOF:
- dcsync single
- adwssearch
# Or allowing the entire category
ADCS-BOF: all
If a category is commented out or the config is left entirely commented out — no BOF tools will be loaded.
Built-in Tools (no BOF required)
- agent_info
- change_directory
- copy_file
- download_file
- execute_powershell
- execute_raw
- execute_shell
- get_downloaded_file
- get_uid
- get_working_directory
- jobs_kill
- jobs_list
- kill_agent
- kill_process
- list_agents
- list_credentials
- adaptix__list_directory
- list_disks
- list_downloads
- list_listeners
- list_processes
- list_targets
- list_task_history
- list_tunnels
- log_finding
- make_directory
- move_file
- port_forward
- adaptix__read_file
- remove_file
- reverse_port_forward
- run_process
- save_writeup
- set_agent_sleep
- start_socks4
- start_socks5
- stop_tunnel
- tag_agent
- view_session_notes
BOF-based Tools
- bof_adbof
- bof_adcsbof
- bof_credsbof
- bof_elevationbof
- bof_executionbof
- bof_injectionbof
- bof_kerbeusbof
- bof_lateralmovement
- bof_ldapbof
- bof_mssqlbof
- bof_postexbof
- bof_processbof
- bof_relayinformerbof
- bof_salbof
- bof_sarbof
Prompt
1) Local Privilege Escalation
You are an experienced red team operator performing post-exploitation in a controlled lab environment.
MCP server are already connected and available to you.
You must use these MCP capabilities to execute commands on the compromised host.
Key Architecture Fact:
Each bof_* tool is a GROUP containing many subcommands, not a single action. Always read the tool's full description to see ALL available subcommands before deciding what to run.
Operational Logic:
NEVER skip recon. Always follow a logical sequence:
1. Establish current context (who you are, network position, system details)
2. Determine whether the host is domain-joined using available system information
3. If the host is domain-joined:
- Perform domain-aware enumeration where relevant
4. If the host is NOT domain-joined:
- Completely ignore domain-related actions
5. Identify privilege escalation vectors through local enumeration
6. Escalate privileges → continue enumeration
Critical Rules:
- Never assume environment context (domain vs non-domain) without verifying it first
- All argument values must come from prior recon, not guesses
- Validate each step before proceeding
- When unsure which subcommand to use, read the full tool description
You already have access to a compromised Windows host.
Your objective:
- Escalate privileges to local administrator on the current machine
IMPORTANT:
- After successful privilege escalation, you MUST spawn a new agent with elevated (administrator/SYSTEM) privileges
Path to agent:
C:\Users\Username\Downloads\agent.exe
Focus only on local privilege escalation to gain privileged access.
⚠️ Disclaimer
This project is intended for authorized security testing and research purposes only.
Do not use this software on systems you do not own or have explicit permission to test.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。