Divine Dragon is my bachelor's degree project. It's more like a pet project, but I have tried to move it closer to the side where it becomes a real product. So, let's explore some features that the Divine Dragon is offering to you.
Divine Dragon has a modular architecture (similar to Metasploit). Implement new modules easily.
Contributors are welcome!
As many tools have it, I decided to implement it here too.
The tool can perform the following basic steps in the enumeration process for pentesters:
- Enumeration of users via Kerberos service (borrowed some code from kerbrute).
- Enumeration of users, groups, computers, DCs, OUs, built-in groups, and Kerberoasting and Asreproasting users via the LDAP service.
- Enumeration of SMB shares, including dumping all files from them if you have sufficient permissions.
The tool allows you to perform the following attacks:
- ASREPRoasting
- Kerberoasting
- Password Spraying via Kerberos and SMB
Inside the tool, you can find the functionality to start the C2 server:
- The server is implemented as a simple web server with an API.
- The server will print your password to authenticate in its API (if you want that).
- The API implements role-based authorization.
The previous feature was the C2 server. As you can assume, there's also an implant. That's basically it. You have my implant, and I call it "agent". An agent is a payload that you can generate inside the CLI of the dragon. The agent is naked for now. It means that it has no evasion, no obfuscation, just the simple logic of the "agent". The agent talks to the C2 and asks it about new jobs. If it finds something in the list of available jobs, it will try to execute it if there is some payload available for the job, of course.
List of available payloads to execute through the agent:
- Enumeration via PowerShell script in the payload that uses PowerView.ps1.
- Exploit for lateral movement/privesc: Pass-the-Hash, Pass-the-Ticket.
- Post-exploitation moment like DCSync attack.
All of the payloads described above are created through a basic "payload generator" module.
- gin - as a main framework for the C2 based on HTTP with REST API.
- gin-jwt - as a middleware for JWT in REST API.
- gokrb5 - to work with Kerberos service.
- go-ldap - to work with LDAP service.
- go-smb2 - to work with SMB service.
- kerbrute - have taken some code from here.
General commands:
list modules- to see the list of implemented or active modules.show info <module>- to see information about a specific module.use <module>- to choose a module.sessions- to check the sessions of the active agents.jobs <agent-uuid>- to check the jobs of a specific agent.logs <agent-uuid>- to check the logs of a specific agent.exit
Module commands:
info- to see information about the chosen module.show options- to see the available options for the chosen module.set <OPTION> <VALUE>- to set a value to the module.run- to execute.back- to return to general commands.
C2 server implemented as an HTTP server with REST API. Currently, certificates for TLS server authentication are generated as self-signed.
The REST API has two roles implemented:
Agent(like an implant, agent, or beacon)Operator(like a Red Team Operator)
Both roles will obtain a Json-Web-Token (JWT) after connecting to C2. Each of the generated tokens is signed with a very large secret. These JWT tokens are active for about 3 hours for each role. It is allowed to refresh the JWT token in 6 hours.
To start a C2, use the c2 module as shown below:
/connect- a route to register a new agent and connect it to C2./agent/jobs- a route to check all jobs of the specific agent./agent/jobs/:job-uuid/payload/- a route for the agent to take a payload for a specific job with UUID./agent/jobs/update- a route to update the status of the job./agent/logs/add- a route to add logs with information about how the job goes.
The operator role has the following API routes:
/login- a route to register a new operator and login at the same time. For now, it is what it is./operator/agents/- a route to check all active agents./operator/agents/:agent-uuid/jobs- a route to check the jobs of a specific agent./operator/agents:agent-uuid/jobs/:job-uuid/status- a route to get the status of a specific job for a specific agent./operator/agents/:agent-uuid/logs- a route to get all logs about the jobs of the specific agent.
/helpers/:job-uuid/:helper-filename- a route to download some helpers for payloads like Mimikatz, PowerView, etc.
Agent is just a small program that will execute whatever it receives from the C2. At the start of the process with the agent binary, it will gather the hostname and username and generate a unique UUID for the agent. After that, it will register with this information on C2.
The agent doesn't implement living-off-the-land techniques. It's more like a proof-of-concept or an example-like program in Go for the purposes of a pentest. So, each payload that you want it to execute will touch the filesystem.
To generate an agent, you have to use a special module called payload_generator/agent. More about it will be below.
The next scheme will summarize the above:
The payload generator is a module of the Divine Dragon. It has two main functionalities:
- Generate a stager (the agent above).
payload_generator/agentmodule in CLI. - Generate a payload for the stager.
payload_generator/shellor any oflocal_<whatever>/<whatever>modules.
For example, to generate a new agent payload, use the payload_generator/agent module as below:
Let's see a scheme with Pass-The-Hash technique that will explain how the Divine Dragon runs exploits locally through the agent:
As you can see, it's all about different executables, which, as I said before, will touch the filesystem.
The Divine Dragon will use module payload_generator/stage_two which is hidden for users in CLI. It will use it to generate all payloads for the agent and supply them to C2.
A little demonstration is below.
Part 1:
Part 2:
The information in this repository is for research and educational purposes and not meant to be used in production environments and/or as part of commercial products.
Whether you want to report a bug, send a patch, or give some suggestions on this package, reach out to me at https://t.me/whatsyourask.