Beelzebub API v1

Overview

The Beelzebub Honeypot Framework provides a flexible system for creating and deploying various types of honeypots that simulate vulnerable services to detect and analyze potential attacks. The framework supports multiple protocols including HTTP, SSH, and TCP, with customizable response behaviors.

API Structure

All API configurations in the Beelzebub Honeypot Framework follow a common pattern:

• apiVersion: Specifies the API version (currently "v1")

• protocol: Defines the protocol being emulated (http, ssh, tcp)

• address: The network address and port to listen on (e.g., ":8080", ":22")

• description: A human-readable description of the honeypot service

Protocol-Specific Configurations

HTTP Honeypot

HTTP honeypots simulate web servers and web applications, allowing for customized responses to incoming HTTP requests.

Sample Configuration:

apiVersion: "v1"
protocol: "http"
address: ":8080"
description: "Apache 401"
commands:
  - regex: ".*"
    handler: "Unauthorized"
    headers:
      - "www-Authenticate: Basic"
      - "server: Apache"
    statusCode: 401

Key Components:

• commands: Array of command configurations

  • regex: Regular expression pattern to match incoming requests

  • handler: Response content to return

  • headers: HTTP headers to include in the response

  • statusCode: HTTP status code to return

Plugin Support:

• The HTTP protocol supports the LLMHoneypot plugin for AI-powered responses

SSH Honeypot

SSH honeypots simulate SSH servers, providing interactive command-line interfaces to attackers.

Sample Configuration (Standard):

apiVersion: "v1"
protocol: "ssh"
address: ":22"
description: "SSH interactive"
commands:
  - regex: "^ls$"
    handler: "Documents Images Desktop Downloads .m2 .kube .ssh .docker"
  # Additional command patterns...
  - regex: "^(.+)$"
    handler: "command not found"
serverVersion: "OpenSSH"
serverName: "ubuntu"
passwordRegex: "^(root|toor)$"
deadlineTimeoutSeconds: 60

Sample Configuration (LLM-powered):

apiVersion: "v1"
protocol: "ssh"
address: ":2222"
description: "SSH interactive ChatGPT"
commands:
  - regex: "^(.+)$"
    plugin: "LLMHoneypot"
serverVersion: "OpenSSH"
serverName: "ubuntu"
passwordRegex: "^(root|qwerty|Smoker666|123456|jenkins|minecraft|sinus|alex|postgres|Ly123456)$"
deadlineTimeoutSeconds: 120
plugin:
    llmProvider: "openai"
    llmModel: "gpt-4o"
    openAISecretKey: "sk-proj-1234567890"

Key Components:

• commands: Array of command patterns and responses

  • regex: Regular expression to match user commands

  • handler: Text output to display in response to the command

  • plugin: Optional plugin name to handle the command (e.g., "LLMHoneypot")

• serverVersion: SSH server version string to display

• serverName: Server name to display (e.g., "ubuntu")

• passwordRegex: Regular expression defining accepted passwords

• deadlineTimeoutSeconds: Session timeout in seconds

Plugin Configuration:

• For LLM-powered SSH honeypots:

  • llmProvider: The LLM service provider (e.g., "openai", "ollama")

  • llmModel: The specific model to use (e.g., "gpt-4o")

  • openAISecretKey: API key for the LLM service

TCP Honeypot

TCP honeypots emulate various TCP-based services with customizable banners.

Sample Configuration:

apiVersion: "v1"
protocol: "tcp"
address: ":3306"
description: "Mysql 8.0.29"
banner: "8.0.29"
deadlineTimeoutSeconds: 10

Key Components:

• banner: Text sent to clients upon connection

• deadlineTimeoutSeconds: Connection timeout in seconds

LLMHoneypot Plugin

The LLMHoneypot plugin provides AI-powered responses to attacker inputs using language models.

Compatibility: Only available for HTTP and SSH protocols.

Configuration Parameters:

• llmProvider: The AI service provider (currently supports "openai", "ollama")

• llmModel: The language model to use (e.g., "gpt-4o")

• openAISecretKey: Authentication key for the LLM service

• prompt: Custom prompt to customize the honeypot

Best Practices

1. Port Selection:

   • Use standard ports for better attacker engagement (e.g., :22 for SSH, :80/:443 for HTTP)

   • For multiple instances of the same protocol, use non-standard ports for additional honeypots

2. Response Configuration:

   • Create realistic command responses that mimic actual systems

   • Include deliberate vulnerabilities or information leaks to engage attackers

3. LLM Integration:

   • Use LLM-powered honeypots for more dynamic and convincing interactions

   • Configure appropriate timeouts to manage resource usage

4. Password Complexity:

   • Include common weak passwords in passwordRegex to attract brute force attempts

   • Mix simple passwords with moderately complex ones for realistic representation

5. Session Management:

   • Set appropriate deadlineTimeoutSeconds based on expected interaction patterns

   • Lower timeouts (10-30 seconds) for simple services

   • Higher timeouts (60-120+ seconds) for interactive sessions

Implementation Examples

Basic HTTP Authentication Honeypot

apiVersion: "v1"
protocol: "http"
address: ":80"
description: "Apache Basic Auth"
commands:
  - regex: ".*"
    handler: "Unauthorized"
    headers:
      - "www-Authenticate: Basic"
      - "server: Apache/2.4.41 (Ubuntu)"
    statusCode: 401

Interactive SSH Honeypot with LLM

apiVersion: "v1"
protocol: "ssh"
address: ":22"
description: "SSH interactive with AI"
commands:
  - regex: "^(.+)$"
    plugin: "LLMHoneypot"
serverVersion: "OpenSSH_8.2p1"
serverName: "ubuntu"
passwordRegex: "^(root|admin|password|123456)$"
deadlineTimeoutSeconds: 120
plugin:
    llmProvider: "openai"
    llmModel: "gpt-4o"
    openAISecretKey: "sk-proj-XXXXXXXXXXXX"

Database Service Honeypot

apiVersion: "v1"
protocol: "tcp"
address: ":3306"
description: "MySQL Server"
banner: "5.7.38-log MySQL Community Server (GPL)"
deadlineTimeoutSeconds: 15