蜂蜜MCP

Install

pip install honeymcp
honeymcp init  # Creates config files

1. Honeypot Deployment

HoneyMCP injects deceptive security-sensitive tools that appear alongside legitimate tools:

Two Modes:

Dynamic Mode (Default) - LLM analyzes your server context and generates domain-specific honeypots: - File server → bypass_file_permissions, read_system_credentials - Database server → dump_admin_credentials, bypass_query_restrictions - API gateway → list_internal_api_keys, access_admin_endpoints

Static Mode - Pre-configured generic honeypots: - list_cloud_secrets, execute_shell_command, read_private_files

Quick Setup with CLI

The easiest way to configure HoneyMCP: ```bash honeymcp init # Creates honeymcp.yaml + .env.honeymcp

LLM Setup (Dynamic Ghost Tools)

Dynamic ghost tools require LLM credentials. Run honeymcp init to generate .env.honeymcp, then add your credentials:

Add to .env.honeymcp:

LLM_PROVIDER=openai
LLM_MODEL=gpt-4o-mini
OPENAI_API_KEY=your_key_here

Supported providers: - LLM_PROVIDER=openai: Requires OPENAI_API_KEY - LLM_PROVIDER=watsonx: Requires WATSONX_URL, WATSONX_APIKEY, WATSONX_PROJECT_ID - LLM_PROVIDER=ollama: Requires OLLAMA_API_BASE (default: http://localhost:11434)

HoneyMCP loads .env.honeymcp first, then falls back to .env. This keeps HoneyMCP credentials separate from your project's environment.

🚀 Quick Start

Basic Usage

Add HoneyMCP to your FastMCP server with one line:

```python from fastmcp import FastMCP from honeymcp import honeypot

mcp = FastMCP("My Server")

@mcp.tool() def my_real_tool(data: str) -> str: """Your legitimate tool""" return f"Processed: {data}"

Usage

honeymcp create-tool "dump container registry credentials"

ToolGen automatically: - Determines tool category (exfiltration, bypass, privilege escalation) - Infers threat level from description keywords - Extracts parameters and types - Generates realistic response templates - Adds tool to both ghost_tools.py and middleware.py - Validates all generated code

Example

$ honeymcp create-tool "list terraform state files with secrets"

✅ Tool created: list_terraform_state
   Category: exfiltration
   Threat Level: critical
   
📝 Agent Reasoning:
   - Analyzing tool description to extract specifications
   - Generating response generator function
   - Validating generated response function
   - Checking code quality and security

The new tool is immediately available in your honeypot catalog.

---

Try the Demo

git clone https://github.com/barvhaim/HoneyMCP.git
cd HoneyMCP
uv sync

Static ghost tools demo:

MCP_TRANSPORT=sse uv run python examples/demo_server.py

Dynamic ghost tools demo (requires LLM credentials in .env.honeymcp):

MCP_TRANSPORT=sse uv run python examples/demo_server_dynamic.py

🔧 Configuration

Optional: remove all persisted attack event files

honeymcp clean-data ```

YAML Config

```yaml

Environment Overrides

HoneyMCP also supports environment overrides:

• HONEYMCP_EVENT_PATH - overrides the base event storage directory

Full Configuration

from pathlib import Path
from honeymcp import honeypot, ProtectionMode

mcp = honeypot(
    mcp,
    # Dynamic ghost tools (default)
    use_dynamic_tools=True,           # LLM-generated domain-specific tools
    num_dynamic_tools=3,              # Number of dynamic tools to generate
    fallback_to_static=True,          # Use static tools if LLM fails

    # Static ghost tools (optional)
    ghost_tools=["list_cloud_secrets", "execute_shell_command"],

    # Protection mode (default: SCANNER)
    protection_mode=ProtectionMode.SCANNER,  # or ProtectionMode.COGNITIVE

    # Other settings
    event_storage_path=Path.home() / ".honeymcp" / "events",
    enable_dashboard=True,
)

Dynamic vs Static Tools: - Dynamic (default): LLM analyzes your server and generates relevant honeypots (requires LLM credentials in .env.honeymcp) - Static: Pre-defined generic tools (no LLM required, set use_dynamic_tools=False)

---

1. Configure Claude Desktop

For stdio transport (recommended - works with all Claude Desktop versions):

Edit ~/Library/Application Support/Claude/claude_desktop_config.json:

{
    "mcpServers": {
      "honeymcp-demo": {
        "command": "uv",
        "args": ["run", "python", "/path/to/HoneyMCP/examples/demo_server.py"],
        "env": {"MCP_TRANSPORT": "stdio"}
      }
    }
}

If your client does not support an env block, launch the server with MCP_TRANSPORT=stdio in your shell.

For Streamable HTTP transport (requires Claude Pro/Max/Team/Enterprise):

1. Start the server:

   MCP_TRANSPORT=http uv run python examples/demo_server.py
   

2. Configure Claude Desktop:

   {
     "mcpServers": {
       "honeymcp-demo": {
         "url": "http://localhost:8000/mcp"
       }
     }
   }