Nebula is a Cloud and (hopefully) DevOps Penetration Testing framework. It is build with modules for each provider and each functionality. As of April 2021, it only covers AWS, but is currently an ongoing project and hopefully will continue to grow to test GCP, Azure, Kubernetes, Docker, or automation engines like Ansible, Terraform, Chef, etc. I started writing it while I was reading "Hands-On AWS Penetration Testing with Kali Linux" (https://www.amazon.com/Hands-Penetration-Testing-Kali-Linux/dp/1789136725) and was based on Pacu (https://github.com/RhinoSecurityLabs/pacu)
Presentations:
- BlackHat Europe 2021: https://www.blackhat.com/eu-21/arsenal/schedule/index.html#nebula-a-case-study-in-penetrating-something-as-soft-as-a-cloud-25174
Currently covers:
- AWS, Azure (Graph and Management API) and DigitalOcean enumeration, exploitation and post-exploitation
There are currently 72 modules covering:
- Reconnaissance
- Enumeration
- Exploit
- Cleanup
Version 3.0 Includes:
- Team cooperation with the client-teamserver architecture
- All the requests requires authentication (except for the authentication request ofc)
- All the information is stored into a MongoDB Server and accessible using commands. The information will ofc have to been enumerated before, but this allows you to not enumerate a certain object
Nebula is coded in python3.11. It uses boto3 library to access AWS. To install, create a venv and install python 3.11+ and install libraries required from requirements.txt
python3 -m venv ./venv
source venv/bin/activate
python3 -m pip install -r requirements.txt
Then install session-manager-plugin. This is needed for SSM modules:
curl "https://s3.amazonaws.com/session-manager-downloads/plugin/latest/ubuntu_64bit/session-manager-plugin.deb" -o "session-manager-plugin.deb"
dpkg -i session-manager-plugin.deb
On windows devices, since less is not installed, I got one from https://github.com/jftuga/less-Windows The prebuilt binary is saved on directory less_binary. Just add that directory to the PATH environment variable and it will be ok.
Then just run teamserver
python3 teamserver.py -dn <workspace name> -p <password>
Same with client client. You will need to create another venv outside of the scope of the teamserver one. Create the venv and install the libraries:
cd client
python3 -m venv ./venv
source venv/bin/activate
python3 -m pip install -r requirements.txt
nebula -w <database name> --password <password> -ah <server host>
...........
...''''''''''''''...
..'''''...........''''''............
..''''.. ...'''''''''''''''...
..'''.. ..............'''''..
.''''. .;loddool:'. ..''''..
..'''. .;clokXWWMWNKkl;. .''''.
.'''. .',,'.. ';dNMMMMMWKko;. .'''..
.''''. .cx0NWWNX0koc;,'cKMMMMMMMMMWXOo:. .''''....
.'''. .',',:oONMMMMMWNNNWMMMMMMWKk0WMMWXx' .''''''''...
..'''. .,dXMMMMMMMMMMMMMNOl',oONWWd. .......'''''..
...'''''.. :o' cXMMMMMMMMMMMMMWNXKKXNWWKxc,. ..''''..
..''''.... oNKl'. ..oXMMMMMMMMMMMMMMMMMMMMMMMMMNKOdc,.. ..''''.
..''''.. ,OWWX0O0XWMMMMMMMMMMMMMMMMMMWWWWMMMMMMMMMWXOxooxk:. ..'''.
..'''''''''''''''''''''. .l0NMMMMMMMMMMMMMMMMMMMMN0dc;;;coONMMMMMMMMMMMMMK: ..'''.
....................... .,dXMMMMMMMMMMMMMMMMMMWX0ko:. .;OWMMMMMMMMMMMWx. .'''.
.oWMMMMMMMMMMMMMMWNXXXWMMWKd' .:lccclodOXWMWd. .'''.
,lc' .................. ',. .,OWMMMMMMMMMMMMXx:'...:0WMMMKl. .. .'oKO, .'''.
,0MWx. .''''''''''''''''''. ;OKOOOO0NWMMMMMMMMMMMMNl. .cdoox0XOl;'....... ... .'''.
.;ol' ................... ;kXWMMMMMMMMMMMMMMMMMWx. .:0WNKkdo:. ... .'''.
.................... .:ldxk0XWMMMMMMMMMMMW0o' .';;,. .... ..'''.
;k00000000000000000000x' ..;lkXWMMMMMMMMMWXkc. ..'''.
.lXWWWWWWWWWWWWWWWWWWMMWKl. ;OWMMMMMMMMMMMWKx:. ..''''.
.,,,,,,,,,,,,,,,,,:kNMMW0o,. 'kWMMMMMMMMMMMMMMWKd,. ..''''..
.:ONMMMNKkdlc:::::::::ccldkKWMMMMMMMMMMMMMMMMMMNOl' ...........'''''..
.,oOXWMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMWXkc....''''''''''...
.':ldkO0000000000000000000000000000000000000000Ox:. ........
...........................................
_ _______ ______ _ _______
( ( /|( ____ \( ___ \ |\ /|( \ ( ___ )
| \ ( || ( \/| ( ) )| ) ( || ( | ( ) |
| \ | || (__ | (__/ / | | | || | | (___) |
| (\ \) || __) | __ ( | | | || | | ___ |
| | \ || ( | ( \ \ | | | || | | ( ) |
| ) \ || (____/\| )___) )| (___) || (____/\| ) ( |
|/ )_)(_______/|/ \___/ (_______)(_______/|/ \|
Because Clouds are so AWSome
-------------------------------------------------------------
Created by: gl4ssesbo1
-------------------------------------------------------------
87 aws 0 gcp 1 azure 0 office365
0 docker 0 kubernetes 3 misc 2 azuread
-------------------------------------------------------------
93 modules 3 cleanup 0 detection
62 enum 11 exploit 1 persistence
1 listeners 0 lateral movement 2 detection bypass
0 privesc 9 reconnaissance 1 stager
3 misc
Remember:
-------------------------------------------------------------
1) Only use this tool if you have permissions from the
infrastructure's owner. Don't be a dick. Don't choose jail.
And if you have some scruples, don't hack others just because
you can (or cannot, in which case that's why you chose this
tool to do it).
2) There is a template file on module directory that you can
use if you want to develop new modules. If you want to
contribute on this tool, be my guest.
3) Thank you for using this tool and Hack the Planet Legally!
-------------------------------------------------------------
[*] Importing sessions found on ~/.aws
[*] Imported sessions found on ~/.aws. Enter 'show credentials' to get the credentials.
(work5)()(Nebula) >>>
Running help command, will give you a list of the commands that can be used:
()()(AWS) >>> help
Help Command: Description:
------------- ------------
help Show help for all the commands
help credentials Show help for credentials
help module Show help for modules
help workspace Show help for credentials
help user-agent Show help for credentials
help shell Show help for shell connections
Module Commands Description
--------------- -----------
show modules List all the modules
show enum List all Enumeration modules
show exploit List all Exploit modules
show persistence List all Persistence modules
show privesc List all Privilege Escalation modules
show reconnaissance List all Reconnaissance modules
show listener List all Reconnaissance modules
show cleanup List all Enumeration modules
show detection List all Exploit modules
show detectionbypass List all Persistence modules
show lateralmovement List all Privilege Escalation modules
show stager List all Reconnaissance modules
use module <module> Use a module.
options Show options of a module you have selected.
run Run a module you have selected. Eg: 'run <module name>'
search Search for a module via pattern. Eg: 'search s3'
back Unselect a module
set <option> Set option of a module. Need to have the module used first.
unset <option> Unset option of a module. Need to have the module used first.
User-Agent commands Description
------------------- -----------
set user-agent windows Set a windows client user agent
set user-agent linux Set a linux client user agent
set user-agent custom Set a custom client user agent
show user-agent Show the current user-agent
unset user-agent Use the user agent that boto3 produces
Workspace Commands Description
------------------ -----------
create workspace <wp> Create a workspace
use workspace <wp> Use one of the workspaces
remove workspace <wp> Remove a workspace
Shell commands Description
------------------- -----------
shell check_env Check the environment you are in, get data and meta-data
shell exit Kill a connection
shell <command> Run a command on a system. You don't need " on the command, just shell <command1> <command2>
When you have a set of credentials, you can enter getuid to get the user or enum_user_privs to check the Read permission of a set of credentials.
(test)()(AWS) >>> getuid
------------------------------------------------
UserId: A******************Q
------------------------------------------------
UserID: A******************Q
Arn: arn:aws:iam::012345678912:user/user_user
Account: 012345678912
[*] Output is saved to './workspaces/test/12_07_2021_02_22_54_getuid_dev_brian'
If the creds do not have the below privs on himself,
STS:GetUserIdentity
IAM:GetUser
IAM:ListAttachedUserPolicies
IAM:GetPolicy (for all policies)
you will get an error:
[*] An error occurred (AccessDenied) when calling the GetUser operation: User: arn:aws:iam::012345678912:user/user_user is not authorized to perform: iam:GetUser on resource: user user_user
This command checks List and Describe Privileges on a set of credentials.
(test)()(AWS) >>> enum_user_privs
User: user_user
UserID: A******************Q
Arn: arn:aws:iam::012345678912:user/user_user
Account: 012345678912
--------------------------
Service: ec2
--------------------------
[*] Trying the 'Describe' functions:
[*] 'describe_account_attributes' worked!
[*] 'describe_addresses' worked!
[*] 'describe_aggregate_id_format' worked!
[*] 'describe_availability_zones' worked!
[*] 'describe_bundle_tasks' worked!
[*] 'describe_capacity_reservations' worked!
[*] 'describe_client_vpn_endpoints' worked!
[*] 'describe_coip_pools' worked!
[*] 'describe_customer_gateways' worked!
[*] 'describe_dhcp_options' worked!
[*] 'describe_egress_only_internet_gateways' worked!
^C[*] Stopping. It might take a while. Please wait.
[*] Output of the allowed functions is saved to './workspaces/test/12_07_2021_02_24_09_enum_user_privs'
[*] The list of the allowed functions is saved to './workspaces/test/12_07_2021_02_24_09_allowed_functions'
You can list all the modules or specific module:
()()(AWS) >>> show modules
cleanup/aws_iam_delete_access_key Delete access key of a user by providing
it.
cleanup/aws_iam_delete_login_profile Delete access of a user to the Management
Console
enum/aws_ec2_enum_elastic_ips Lists User data of an Instance provided.
Requires Secret Key and Access Key of an IAM that has access
to it.
enum/aws_ec2_enum_images List all ec2 images. Needs credentials of an
IAM with DescribeImages right. Output is dumpled on a file.
It takes a sh*tload of time, unfortunately. And boy, is it a
huge output.
enum/aws_ec2_enum_instances Describes instances attribues: Instances, VCP,
Zones, Images, Security Groups, Snapshots, Subnets, Tags,
Volumes. Requires Secret Key and Access Key of an IAM that
has access to all or any of the API calls:
DescribeAvailabilityZones, DescribeImages,
DescribeInstances, DescribeKeyPairs, DescribeSecurityGroups,
DescribeSnapshots, DescribeSubnets, DescribeTags,
DescribeVolumes, DescribeVpcs
And like that you can use:
show module
show enum
show exploit
show persistence
show privesc
show reconnaissance
show listener
show cleanup
show detection
show detectionbypass
show lateralmovement
show stager
Use search command to search modules with a specific word:
()()(AWS) >>> search instance
enum/aws_ec2_enum_instances Describes instances attribues: Instances, VCP,
Zones, Images, Security Groups, Snapshots, Subnets, Tags,
Volumes. Requires Secret Key and Access Key of an IAM that
has access to all or any of the API calls:
DescribeAvailabilityZones, DescribeImages,
DescribeInstances, DescribeKeyPairs, DescribeSecurityGroups,
DescribeSnapshots, DescribeSubnets, DescribeTags,
DescribeVolumes, DescribeVpcs
enum/aws_iam_list_instance_profiles List all the instance profiles.
exploit/aws_ec2_create_instance_with_user_data You must provide policies in JSON format in
IAM. However, for AWS CloudFormation templates formatted in
YAML, you can provide the policy in JSON or YAML format. AWS
CloudFormation always converts a YAML policy to JSON format
before submitting it to IAM.
()()(AWS) >>>
To use a module, just type use and the name of the module. The 3 brackets will have the name of the module.
(work1)()(enum/aws_ec2_enum_instances) >>> use module enum/aws_iam_get_group
(work1)()(enum/aws_ec2_enum_instances) >>>
Using options, we can list the information on the module:
(work1)()(enum/aws_ec2_enum_instances) >>> options
Desctiption:
-----------------------------
Describes instances attribues: Instances, VCP, Zones, Images, Security Groups, Snapshots, Subnets, Tags, Volumes. Requires Secret Key and Access Key of an IAM that has access to all or any of the API calls: DescribeAvailabilityZones, DescribeImages, DescribeInstances, DescribeKeyPairs, DescribeSecurityGroups, DescribeSnapshots, DescribeSubnets, DescribeTags, DescribeVolumes, DescribeVpcs
Author:
-----------------------------
name: gl4ssesbo1
twitter: https://twitter.com/gl4ssesbo1
github: https://github.com/gl4ssesbo1
blog: https://www.pepperclipp.com/
AWSCLI Command:
-----------------------------
aws ec2 describe-instances --region {} --profile {}
Needs Credentials: True
-----------------------------
Options:
-----------------------------
SERVICE: ec2
Required: true
Description: The service that will be used to run the module. It cannot be changed.
INSTANCE-ID:
Required: false
Description: The ID of the instance you want to enumerate. If not supplied, all instances will be enumerated.
(work1)()(enum/aws_ec2_enum_instances) >>>
To set options, use set and the name of the option:
(work1)()(enum/aws_ec2_enum_instances) >>> set INSTANCE-ID 1234
(work1)()(enum/aws_ec2_enum_instances) >>> options
Desctiption:
-----------------------------
Describes instances attribues: Instances, VCP, Zones, Images, Security Groups, Snapshots, Subnets, Tags, Volumes. Requires Secret Key and Access Key of an IAM that has access to all or any of the API calls: DescribeAvailabilityZones, DescribeImages, DescribeInstances, DescribeKeyPairs, DescribeSecurityGroups, DescribeSnapshots, DescribeSubnets, DescribeTags, DescribeVolumes, DescribeVpcs
Author:
-----------------------------
name: gl4ssesbo1
twitter: https://twitter.com/gl4ssesbo1
github: https://github.com/gl4ssesbo1
blog: https://www.pepperclipp.com/
Needs Credentials: True
-----------------------------
AWSCLI Command:
-----------------------------
aws ec2 describe-instances --region {} --profile {}
Options:
-----------------------------
SERVICE: ec2
Required: true
Description: The service that will be used to run the module. It cannot be changed.
INSTANCE-ID: 1234
Required: false
Description: The ID of the instance you want to enumerate. If not supplied, all instances will be enumerated.
(work1)()(enum/aws_ec2_enum_instances) >>>
Also unsetting them, using unset.
(work1)()(enum/aws_ec2_enum_instances) >>> unset INSTANCE-ID
(work1)()(enum/aws_ec2_enum_instances) >>>
To run the module, if it requires credentials, you will need to have imported a set of credentials with the permission required to run it. This is shown on a module's options as:
Needs Credentials: True
-----------------------------
To run it, just enter run. Depending on the output, it will either show a pagainated view, or just print it. The pagination, uses less binary, which for Windows uses the binary from https://github.com/jftuga/less-Windows. A copy of the exe is on less_binary directory. The output is also saved on files on the workspace directory:
(work1)()(enum/aws_ec2_enum_instances) >>> run
[*] Content dumped on file './workspaces/work1/16_04_2021_18_16_48_ec2_enum_instances'.
####Inputing Credentials Nebula can use both AccessKeyID + SecretKey combination and AccessKeyID + SecretKey+SessionKey combination to authenticate into the infratructure. To insert a set of credentials, use:
()()(AWS) >>> set credentials test1
Profile Name: test1
Access Key ID: A*********2
Secret Key ID: a****************************7
Region: us-west-3
Do you also have a session token?[y/N]
[*] Credentials set. Use 'show credentials' to check them.
[*] Currect credential profile set to 'test1'.Use 'show current-creds' to check them.
And you will get some inputs allowing you to set them. Session token can be added when entering credentials, by inputing y when asked Do you also have a session token?[y/N].
####Using Credentials To use another credential, just enter:
()()(AWS) >>> use credentials test1
[*] Currect credential profile set to 'test1'.Use 'show current-creds' to check them.
####Current Credentials When you enter the credentials, they are automatically made the current credentials, meaning the ones you will authenticate with. To check the current credentials, use:
()()(AWS) >>> show current-creds
{
"profile": "test1",
"access_key_id": "A*********2",
"secret_key": "a****************************7",
"region": "us-west-3"
}
####Removing Credentials In case you don't want your credentials, you can can remove them using:
()()(AWS) >>> remove credentials test1
You are about to remove credential 'test1'. Are you sure? [y/N] y
####Dumping and importing credentials In case you want your credentials saved on the machine, you can use:
()()(AWS) >>> dump credentials
[*] Credentials dumped on file './credentials/16_04_2021_17_37_59'.
And they will be saved on a file containing the time and date of the dump on directory credentials on Nebula directory. To import them, just enter:
()()(AWS) >>> import credentials 16_04_2021_17_37_59
()()(AWS) >>> show credentials
[
{
"profile": "test1",
"access_key_id": "A*********2",
"secret_key": "a****************************7",
"region": "us-west-3"
}
]
Nebula uses workspaces to save the output from every command. The output is saved as json data (except for s3_name_fuzzer which saves it as XML) on a folder created on directory workspaces.
To create one, enter:
()()(AWS) >>> create workspace work1
[*] Workspace 'work1' created.
[*] Current workspace set at 'work1'.
(work1)()(AWS) >>> ls ./workspaces
Directory: C:\Users\***\Desktop\Nebula\workspaces
Mode LastWriteTime Length Name
---- ------------- ------ ----
d----- 4/16/2021 5:42 PM work1
-a---- 4/16/2021 4:40 PM 0 __init__.py
When created, the first brackets will contain the name of the workspace you are working at. If you want to use an existing workspace, just type:
()()(AWS) >>> use workspace work1
(work1)()(AWS) >>>
Workspaces are required to be used, so even if you are not using any at the moment, while running a module, it will ask you to create one with random name or to just create one with a custom name yourself.
()()(enum/aws_ec2_enum_instances) >>> run
A workspace is not configured. Workstation 'qxryiuct' will be created. Are you sure? [y/N] n
[*] Create a workstation first using 'create workstation <workstation name>'.
()()(enum/aws_ec2_enum_instances) >>>
To get a list of workspaces, use:
(work1)()(enum/aws_ec2_enum_instances) >>> show workspaces
-----------------------------------
Workspaces:
-----------------------------------
work1
(work1)()(enum/aws_ec2_enum_instances) >>>
To remove a workspace, enter:
()()(AWS) >>> remove workspace work1
[*] Are you sure you want to delete the workspace? [y/N] y
()()(AWS) >>> show workspaces
-----------------------------------
Workspaces:
-----------------------------------
()()(AWS) >>>
To create a Reverse Shell, you need to create a stager and run a listener. To use this feature, you need to have Nebula run as root (to open ports).
To generate a stager, use modules on stagers:
()()(AWS) >>> use module stager/aws_python_tcp
()()(stager/aws_python_tcp) >>> options
Desctiption:
-----------------------------
The TCP Reverse Shell that is used by listeners/aws_python_tcp_listener
Author:
-----------------------------
name: gl4ssesbo1
twitter: https://twitter.com/gl4ssesbo1
github: https://github.com/gl4ssesbo1
blog: https://www.pepperclipp.com/
Needs Credentials: False
-----------------------------
AWSCLI Command:
-----------------------------
None
Options:
-----------------------------
SERVICE: none
Required: true
Description: The service that will be used to run the module. It cannot be changed.
HOST:
Required: true
Description: The Host/IP of the C2 Server.
PORT:
Required: true
Description: The C2 Server Port.
FORMAT:
Required: true
Description: The format of the stager. Currently only allows 'py' for Python and 'elf' for ELF Binary.
CALLBACK-TIME: None
Required: true
Description: The time in seconds between callbacks from Stager. The Stager calls back even if the server crashes or is stoped in a loop.
OUTPUT-FILE-NAME:
Required: true
Description: The name of the stager output file.
The options to fill are:
- HOST: The IP or domain of the C2 Server
- Port: The C2 Server Port
- Format: Currently only supports python raw file and elf binary
- Callback-Time: The time in seconds for which the sessions should call back. It calls back even if a current session is up, and even if the server crushes or is closed, so that you don't loose access to the machine.
- Output File Name: The name of the output file.
Running the module will generate a stager saved on ./workspaces/workspacename/stagername
The listener is simple. Just configure Host (by default set to 0.0.0.0) and Port and it creates the server. To run the listener, you need to have Nebula run as root.
()()(stager/aws_python_tcp) >>> use module listeners/aws_python_tcp_listener
()()(listeners/aws_python_tcp_listener) >>> options
Desctiption:
-----------------------------
TCP Listener for Reverse Shell stagers/aws_python_tcp
Author:
-----------------------------
name: gl4ssesbo1
twitter: https://twitter.com/gl4ssesbo1
github: https://github.com/gl4ssesbo1
blog: https://www.pepperclipp.com/
Needs Credentials: False
-----------------------------
AWSCLI Command:
-----------------------------
None
Options:
-----------------------------
SERVICE: none
Required: true
Description: The service that will be used to run the module. It cannot be changed.
HOST: 0.0.0.0
Required: true
Description: The Host/IP of the C2 Server.
PORT:
Required: true
Description: The C2 Server Port.
User agents can be set as linux ones, windows ones or custom. To show them, just use show.
()()(AWS) >>> set user-agent linux
User Agent: Boto3/1.9.89 Python/3.8.1 Linux/4.1.2-34-generic was set
()()(AWS) >>> show user-agent
[*] User Agent is: Boto3/1.9.89 Python/3.8.1 Linux/4.1.2-34-generic
()()(AWS) >>> set user-agent windows
User Agent: Boto3/1.7.48 Python/3.9.1 Windows/7 Botocore/1.10.48 was set
()()(AWS) >>> show user-agent
[*] User Agent is: Boto3/1.7.48 Python/3.9.1 Windows/7 Botocore/1.10.48
()()(AWS) >>> set user-agent custom
Enter the User-Agent you want: sth
User Agent: sth was set
()()(AWS) >>> show user-agent
[*] User Agent is: sth
()()(AWS) >>>
To unset a user agent, enter:
()()(AWS) >>> unset user-agent
[*] User Agent set to empty.
Which will have the system's user agent.
