Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "honeymcp": {
      "command": "uvx",
      "args": [
        "honeymcp"
      ]
    }
  }
}

Or install with pip: pip install honeymcp

README Excerpt

<img src="https://github.com/user-attachments/assets/34f18118-1490-4f06-af08-f2efb0ecec79" alt="HoneyMCP logo" width="300" height="300" /> **Detect AI Agent Attacks Through Deception** HoneyMCP is a defensive security tool that adds deception capabilities to Model Context Protocol (MCP) servers. It injects "ghost tools" (fake security-sensitive tools) that act as honeypots, detecting two critical threat categories:

Tools (20)

HONEYMCP_EVENT_PATHOPENAI_API_KEYWATSONX_PROJECT_IDassume_iam_rolebypass_query_restrictionsbypass_security_checkdisable_2fa_requirementdisable_security_filtersdump_database_credentialsdump_ml_model_weightsdump_session_tokensescalate_privilegesexecute_shell_commandexport_audit_logsexport_user_dataget_api_keysinject_system_messagelist_cloud_secretslist_github_tokenslist_kubernetes_secrets

Topics

agentic-aihoneypotmcp-server

HoneyMCP

Quick Install

README Excerpt

Tools (20)

Topics

Related AI/LLM Servers