Maltego TRX Python Library

Release Notes

1.6.1: Update cryptography and Flask dependency and deprecate Python 3.7

1.6.0: Automatically generate am .mtz for your local transforms

1.5.2: Add logging output for invalid / missing params in xml serialization

1.5.1: Add ignored files to starter and use README for pypi

1.5.0: XML Serialization via ElementTree instead of string interpolation

1.4.4: Added skeletons for csv export in template dir and made project.py application import compatible with docs

1.4.0 + 1.4.1: Both versions are incompatible with python3.7 and lower.

1.4.2: Fixed python3.6 incompatibility

Getting Started

Note: Support for Python 2 has been officially discontinued as of July 2021. Please use Python 3.8 or higher to use up-to-date versions of Maltego TRX.

To install the trx library run the following command:

pip install maltego-trx

After installing, you can create a new project by running the following command:

maltego-trx start new_project

This will create a folder new_project with the recommended project structure.

If you want to copy the starter files to your current directory, run the following command:

maltego-trx init

Alternatively, you can copy either the gunicorn or apache example projects from the demo directory. These also include Dockerfile and corresponding docker-compose configuration files for production deployment.

Adding a Transform:

Add a new transform by creating a new python file in the "transforms" folder of your directory.

Any file in the folder where the class name matches the filename, and the class inherits from Transform, will automatically be discovered and added to your server.

A simple transform would look like the following:

new_project/transforms/GreetPerson.py

from maltego_trx.entities import Phrase
from maltego_trx.transform import DiscoverableTransform


class GreetPerson(DiscoverableTransform):
    """
    Returns a phrase greeting a person on the graph.
    """

    @classmethod
    def create_entities(cls, request, response):
        person_name = request.Value

        response.addEntity(Phrase, "Hi %s, nice to meet you!" % person_name)

Running The Transform Server

For Development

You can start the development server, by running the following command:

python project.py runserver

This will start up a development server that automatically reloads every time the code is changed.

For Production

You can run a gunicorn transform server, after installing gunicorn on the host machine and then running the command:

gunicorn --bind=0.0.0.0:8080 --threads=25 --workers=2 project:application

For publicly accessible servers, it is recommended to run your Gunicorn server behind proxy servers such as Nginx.

Run a Docker Transform server

The demo folder provides an example project. The Docker files given can be used to set up and run your project in Docker.

The Dockerfile and docker-compose file can be used to easily set up and run a development transform server.

If you have copied the docker-compose.yml, Dockerfile and prod.yml files into your project, then you can use the following commands to run the server in Docker.

Run the following to start the development server:

docker-compose up

Run the following command to run a production gunicorn server:

docker-compose -f prod.yml up --build

For publicly accessible servers, it is recommended to run your Gunicorn server behind proxy servers such as Nginx.

Local Transforms

Documentation

Transforms written using this library can be used as either local or server transforms.

To run a local transform from your project, you will need to pass the following arguments:

project.py local <transform_name>

You can find the correct transform_name to use by running python project.py list.

Caveats

The following values are not passed to local transforms, and will have dummy values in their place:

• type: local.Unknown
• weight: 100
• slider: 100
• transformSettings: {}

Using the Transform Registry

Added in 1.4.0 (July 2021)

The Transform Registry enables you to annotate Transforms with metadata like display name, description, input and output entities as well as settings. The Transform Registry will automatically generate CSV files that you can import into the pTDS and/or your iTDS.

Configuring the Registry

You can configure your registry with all the info you would normally add for every transform/seed on a TDS. We recommend creating your registry in an extra file, traditionally called extensions.py, to avoid circular imports.

# extensions.py
from maltego_trx.decorator_registry import TransformRegistry

registry = TransformRegistry(
    owner="ACME Corporation",
    author="John Doe <[email protected]>",
    host_url="https://transforms.acme.org",
    seed_ids=["demo"]
)

# The rest of these attributes are optional

# metadata
registry.version = "0.1"

# transform suffix to indicate datasource
registry.display_name_suffix = " [ACME]"

# reference OAuth settings
registry.oauth_settings_id = ['github-oauth']

Annotating Transforms

# transforms/GreetPerson.py
...
from extensions import registry


@registry.register_transform(
    display_name='Greet Person',
    input_entity='maltego.Phrase',
    description='Returns a phrase greeting a person on the graph.',
    output_entities=['maltego.Phrase'],
    disclaimer='This disclaimer is optional and has to be accepted before this transform is run'
)
class GreetPerson(DiscoverableTransform):

    @classmethod
    def create_entities(cls, request, response):
        ...

Pro Tip: If the display_name is either None or "", the registry will try to create a display name from the class name:

• DNSToIP 'DNS To IP'
• GreetPerson 'Greet Person'

Transform Settings

You can declare transform settings in a central location and add them to the registry.

Configuring Global Settings

These settings will apply to all transforms which can be very helpful for api keys.

# settings.py
from maltego_trx.decorator_registry import TransformSetting

api_key_setting = TransformSetting(name='api_key',
                                   display_name='API Key',
                                   setting_type='string',
                                   global_setting=True)

# extensions.py
from settings import api_key_setting

from maltego_trx.decorator_registry import TransformRegistry

registry = TransformRegistry(
    owner="ACME Corporation",
    author="John Doe <[email protected]>",
    host_url="https://transforms.acme.org",
    seed_ids=["demo"]
)

registry.global_settings = [api_key_setting]

Configuring Settings per Transform

Settings that aren't required for every transform have to be added to the register_transform decorator explicitly. To access the setting on the request, use the id property, which will have the global prefix if it's a global setting. The name property won't work on global settings.

# settings.py
...

language_setting = TransformSetting(name='language',
                                    display_name="Language",
                                    setting_type='string',
                                    default_value='en',
                                    optional=True,
                                    popup=True)

# transforms/GreetPerson.py
...
from settings import language_setting

from maltego_trx.transform import DiscoverableTransform


@registry.register_transform(display_name="Greet Person",
                             input_entity="maltego.Phrase",
                             description='Returns a phrase greeting a person on the graph.',
                             settings=[language_setting])
class GreetPerson(DiscoverableTransform):

    @classmethod
    def create_entities(cls, request: MaltegoMsg, response: MaltegoTransform):
        language = request.getTransformSetting(language_setting.id)
        ...

Exporting the TDS Configuration

To export the configurations, use the registry methods write_transforms_config() and write_settings_config(). These methods have to executed after they have been registered with the TRX server.

# project.py

import sys
import transforms

from maltego_trx.registry import register_transform_function, register_transform_classes
from maltego_trx.server import application
from maltego_trx.handler import handle_run

# register_transform_function(transform_func)
from extensions import registry

register_transform_classes(transforms)

registry.write_transforms_config()
registry.write_settings_config()

handle_run(__name__, sys.argv, application)

Generating an .mtz config with your local Transforms

Since maltego-trx>=1.6.0 you can generate an .mtz config file with your local transforms.

If you're already using the TransformRegistry, just invoke the write_local_config() method.

# project.py

registry.write_local_mtz()

This will create a file called local.mtz in the current directory. You can then import this file into Maltego and start using your local transforms faster. Just remember that settings are not passed to local transforms.

The method takes in the same arguments as the interface in the Maltego client. If you are using a virtualenv environment, you might want to change the command argument to use that.

# project.py

registry.write_local_mtz(
    mtz_path: str = "./local.mtz", # path to the local .mtz file
    working_dir: str = ".",
    command: str = "python3", # for a venv you might want to use `./venv/bin/python3`
    params: str = "project.py",
    debug: bool = True
)

Legacy Transforms

Documentation

If you have old TRX transforms that are written as functions, they can be registered with the server using the maltego_trx.registry.register_transform_function method.

In order to port your old transforms, make two changes:

1. Import the MaltegoTransform class from the maltego_trx package instead of from a local file.
2. Call the register_transform_function in order for the transform to be registered in your project.

For example

In the legacy transform file, change:

from Maltego import *

def old_transform(m):

To:

from maltego_trx.maltego import MaltegoTransform


def old_transform(m):
    ...

In the project.py file add the following:

from maltego_trx.registry import register_transform_function
from legacy_transform import trx_DNS2IP

register_transform_function(trx_DNS2IP)

CLI

The following commands can be run using the project.py file.

Run Server

python project.py runserver

Start a development server that you can use to develop new transforms.

List

python project.py list

List the available transforms together with their transform server URLs and local transform names.

Reference

Constants

The following constants can be imported from maltego_trx.maltego.

Message Types:

• UIM_FATAL
• UIM_PARTIAL
• UIM_INFORM
• UIM_DEBUG

Please take note: You need to enable the debug filter option in the Desktop client Output window to view debug transform messages.

Bookmark Colors:

• BOOKMARK_COLOR_NONE
• BOOKMARK_COLOR_BLUE
• BOOKMARK_COLOR_GREEN
• BOOKMARK_COLOR_YELLOW
• BOOKMARK_COLOR_PURPLE
• BOOKMARK_COLOR_RED

Link Styles:

• LINK_STYLE_NORMAL
• LINK_STYLE_DASHED
• LINK_STYLE_DOTTED
• LINK_STYLE_DASHDOT

Enums

Overlays:

Overlays Enums are imported from maltego_trx.overlays

Overlay OverlayPosition:

• NORTH = "N"
• SOUTH = "S"
• WEST = "W"
• NORTH_WEST = "NW"
• SOUTH_WEST = "SW"
• CENTER = "C"

Overlay Type

• IMAGE = "image"
• COLOUR = "colour"
• TEXT = "text"

Request/MaltegoMsg

The request/maltego msg object given to the transform contains the information about the input entity.

Attributes:

• Value: str: The display value of the input entity on the graph
• Weight: int: The weight of the input entity
• Slider: int: Results slider setting in the client
• Type: str: The input entity type
• Properties: dict(str: str): A key-value dictionary of the input entity properties
• TransformSettings: dict(str: str): A key-value dictionary of the transform settings
• Genealogy: list(dict(str: str)): A key-value dictionary of the Entity genealogy, this is only applicable for extended entities e.g. Website Entity

Methods:

• getProperty(name: str): Get a property value of the input entity
• getTransformSetting(name: str): Get a transform setting value
• clearLegacyProperties(): Delete (duplicate) legacy properties from the input entity. This will not result in property information being lost, it will simply clear out some properties that the TRX library duplicates on all incoming Transform requests. In older versions of TRX, these Entity properties would have a different internal ID when sent the server than what the Maltego client would advertise in the Entity Manager UI. For a list of Entities with such properties and their corresponding legacy and actual IDs, see entity_property_map in maltego_trx/entities.py. For the majority of projects this distinction can be safely ignored.

Response/MaltegoTransform

Methods:

• addEntity(type: str, value: str) -> Entity: Add an entity to the transform response. Returns an Entity object created by the method.
• addUIMessage(message: str, messageType='Inform'): Return a UI message to the user. For message type, use a message type constant.

Entity

Methods:

• setType(type: str): Set the entity type (e.g. "Phrase" for maltego.Phrase entity)
• setValue(value: str): Set the entity value
• setWeight(weight: int): Set the entity weight
• addDisplayInformation(content: str, title: str): Add display information for the entity.
• addProperty(fieldName: str, displayName: str, matchingRule: str, value: str): Add a property to the entity. Matching rule can be strict or loose.
• addOverlay(propertyName: str, position: OverlayPosition, overlay_type: OverlayType): Add an overlay to the entity. OverlayPosition and OverlayType are defined in the maltego_trx.overlays

Overlay can be added as Text, Image or Color

        
        person_name = request.Value
        entity = response.addEntity(Phrase, "Hi %s, nice to meet you!" % person_name)

        # Normally, when we create an overlay, we would reference a property name so that Maltego can then use the
        # value of that property to create the overlay. Sometimes that means creating a dynamic property, but usually
        # it's better to either use an existing property, or, if you created the Entity yourself, and only need the
        # property for the overlay, to use a hidden property. Here's an example of using a dynamic property:
        entity.addProperty(
            'dynamic_overlay_icon_name', 
            displayName="Name for overlay image", 
            value="Champion"  # references an icon in the Maltego client
        )
        entity.addOverlay('dynamic_overlay_icon_name', OverlayPosition.WEST, OverlayType.IMAGE)

        # DISCOURAGED:
        # You *can* also directly supply the string value of the property, however this is not recommended. Why? If
        # the entity already has a property of the same ID (in this case, "DE"), then you would in fact be assigning the
        # value of that property, not the string "DE", which is not the intention. Nevertheless, here's an example:
        entity.addOverlay(
            'DE', # name of an icon, however, could also accidentally be a property name
            OverlayPosition.SOUTH_WEST, 
            OverlayType.IMAGE
        )

        # Overlays can also be used to display extra text on an entity:
        entity.addProperty("exampleDynamicPropertyName", "Example Dynamic Property", "loose", "Maltego Overlay Testing")
        entity.addOverlay('exampleDynamicPropertyName', OverlayPosition.NORTH, OverlayType.TEXT)

        # Or a small color indicator:
        entity.addOverlay('#45e06f', OverlayPosition.NORTH_WEST, OverlayType.COLOUR)

• setIconURL(url: str): Set the entity icon URL
• setBookmark(bookmark: int): Set bookmark color index (e.g. -1 for BOOKMARK_COLOR_NONE, 3 for BOOKMARK_COLOR_PURPLE)
• setNote(note: str): Set note content
• setGenealogy(genealogy: dict): Set genealogy

Link Methods:

• setLinkColor(color: str): Set the link color (e.g. hex "#0000FF" for blue)
• setLinkStyle(style: int): Set the link style index (e.g. 0 for LINK_STYLE_NORMAL, 2 for LINK_STYLE_DOTTED)
• setLinkThickness(thick: int): Set link thickness (default is 1)
• setLinkLabel(label: str): Set the label of the link
• reverseLink(): Reverse the link direction
• addCustomLinkProperty(fieldName=None, displayName=None, value=None): Set a custom property for the link