Splunk® SOAR (Cloud)

Develop Apps for Splunk SOAR (Cloud)

The classic playbook editor will be deprecated in early 2025. Convert your classic playbooks to modern mode.
After the future removal of the classic playbook editor, your existing classic playbooks will continue to run, However, you will no longer be able to visualize or modify existing classic playbooks.
For details, see:

Tutorial: Use the app wizard to develop an app framework

This section covers app development on . It discusses steps needed to develop an app that interacts with an external device or service, and implements actions that can be run through .

This tutorial uses the App Wizard to design and develop an app framework. Code is then added to the framework to connect to an external service and return the results back to .

Access the Splunk SOAR OVA

It is best to use a running Splunk SOAR instance for developing a Splunk SOAR app. You can download the OVA by going to https://www.phantom.us/download/ and registering for an account on the portal. Once the account request process is complete and you have created your own login, you can download the OVA from the Splunk SOAR community portal at https://my.phantom.us/.

App overview

An app made for has two basic interfaces:

  • An interface to interact with an external device or service, such as whois or VirusTotal, to implement an action.
  • An interface to interact with to run actions, send progress messages, and return action results.

The app framework created by the App Wizard will contain sample code that does both.

Set up the dev environment

Once the OVA file is downloaded, boot it up in VMware, or your virtual machine software of choice, and set the root and user passwords. Remote root login is disabled, so you will have to SSH in as user and sudo to root.

The firewall is pre-configured to allow SSH connections.

SSH to Splunk SOAR

mymbp:~ alice$ ssh user@172.16.242.150
user@172.16.242.150's password:
[user@localhost ~]$ sudo su -
[sudo] password for user:
[root@localhost ~]#

Set the host name

Optionally, set the host name of the instance:

[root@localhost ~]# hostname appdev
[root@localhost ~]#

User 'phantom'

It's good practice to use a non-root user for app development. The user phantom is already present in the OVA and can be used for this purpose. However, this user does not have a password set and as such can't SSH into the Splunk SOAR instance.

Set the password for the user phantom:

[root@localhost ~]# passwd phantom
Changing password for user phantom.
New password:
Retype new password:
passwd: all authentication tokens updated successfully.
[root@localhost ~]# 

Exit and login as phantom.

[root@localhost ~]# exit
logout
[user@localhost ~]$ exit
logout
Connection to 172.16.242.150 closed.
mymbp:~ alice$ ssh phantom@172.16.242.150
phantom@172.16.242.150's password:
[phantom@appdev ~]$

View the ipinfo.io documentation

Use the ipinfo.io API documentation to complete the implementation of the app. The documentation can be found at https://ipinfo.io/developers. From this documentation, determine what requests to make for implementing your app actions. You want to find a request for resolving a host name and a request for getting geolocation information for an IP. Also, look for a request that will help you implement your test connectivity action, which verifies that your asset configuration is correct.

According to ipinfo.io's documentation, these are most likely the requests you want to send for each action:

Test connectivity can be any endpoint. If possible, use a /version or /ping endpoint if existing, but if not you can do any small request.

Using cURL to test the API

Next, check to see if these requests will all work as expected. cURL is a great tool to use for this.

[phantom@appdev ~]$ curl "http://ipinfo.io/8.8.8.8/geo"
{
  "ip": "8.8.8.8",
  "city": "Mountain View",
  "region": "California",
  "country": "US",
  "loc": "37.3860,-122.0840",
  "postal": "94035",
  "phone": "650"
}
[phantom@appdev ~]$ curl "http://ipinfo.io/8.8.8.8/hostname"
google-public-dns-a.google.com

[phantom@appdev ~]$ curl "http://ipinfo.io/region"
California

[phantom@appdev ~]$

Fortunately, every request returns the necessary information to implement the app's actions within .

App Wizard

For this scenario, use the App Wizard to create an app that integrates with the ipinfo.io service by implementing the geolocate ip action.

  1. Navigate to the Main Menu.
  2. Select Apps.
  3. Click App Wizard.

The App Wizard is broken up in multiple tabs, each representing a logical part of an app. Most of the input elements within the App Wizard have an information icon. On mouse hover, these icons reveal detailed help tips, including naming conventions.

Basic details

For our tutorial app, fill the following values:

Field Value
App Name ipinfo io
App Description This app implements investigative actions on ipinfo.io geolocation service
Product Vendor ipinfo.io
Product Name ipinfo.io
App Publisher Tutorial


You can keep the Distribute with source unchecked.

Click on the logo drop boxes to upload a logo. Ideally, the logos are SVG files, but PNG files are also supported.

Asset configuration

To get geolocation information about an ip from ipinfo.io, connect to https://ipinfo.io/<ip_to_query>. The base URL of https://ipinfo.io can be hardcoded in the app, but for this scenario, add a configuration option.

  1. Click the Add Configuration Option button.
    This results in a single row where you can set your configuration options.
  2. Fill in the following values:
    Field Value
    Name base_url
    Data Type String
    Description The Base URL to connect to
    Required Checked
  3. Click More, set the default value to https://ipinfo.io, and click Update.
    This allows the Splunk SOAR platform to display a default value for users when they configure a new asset.

This is useful when there are generally unchanging values for certain parameters like an SSH port number. In this case, if the ipinfo.io were to change domains but offer the exact same service, you do not need to change any app code.

App actions

A list of recommended actions that the app implements is based on the category you select. For this tutorial, select the following:

  1. Select the App Category of information.
  2. Select the Available Action geolocate ip and move it to the Selected Actions.
  3. Fill in the following values for the test connectivity actions:
    Field Value
    REST Endpoint "/region"
  4. Fill in the following value for the geolocate ip action
    Field Value
    REST Endpoint "/{0}/geo".format(ip)

Custom actions

To add any actions other than the recommended actions, you can configure them here. For this tutorial scenario, add a custom action called lookup ip.

  1. Click Add Action and fill in the following values:
    Field Value
    Action Name lookup ip
    Description Returns the resolved host name
    Action Type Investigate
    Read Only Checked
    REST Endpoint "/{0}/hostname".format(ip)
    Verbose Description Use this action to return only the host name from ipinfo.io
  2. Click Add Parameter and fill in the following values:
    Field Value
    Name ip
    Data type String
    Description IP to get the host name of
    Required Checked
    Primary Checked
  3. Click the Contains check box and type ip in the drop-down box.
  4. Click Update.
  5. Click Submit to save the app TAR file on your local machine.

Compiling and installing the app

Copy the app TAR file to the OVA as user phantom. Replace ipinfoio.tgz with the name of your file.

Copy the app TAR file to the OVA

AMBP2:appdev_tutorial alice$ scp ./ipinfoio.tgz phantom@172.16.242.177:~/
phantom@172.16.242.177's password:
ipinfoio.tgz                                                                             100% 7720     3.6MB/s   00:00
AMBP2:appdev_tutorial alice$

Untar the app source TAR file

Untar the app source TAR file to compile and install the app.

AMBP2:appdev_tutorial alice$ ssh phantom@172.16.242.177
phantom@172.16.242.177's password:
[phantom@phantom ~]$ ls
ipinfoio.tgz
[phantom@phantom ~]$ tar -zxvf ipinfoio.tgz
phipinfoio/
phipinfoio/readme.html
phipinfoio/ipinfoio_consts.py
phipinfoio/ipinfoio.png
phipinfoio/ipinfoio_connector.py
phipinfoio/__init__.py
phipinfoio/exclude_files.txt
phipinfoio/ipinfoio.json

Compile and install the app

You can't install the app from the TAR file that downloaded directly from the App Wizard. It needs to be compiled before you can install the app.

To compile your app, use the phenv compile_app command.

Example

[phantom@phantom ~]$ cd phipinfoio
[phantom@phantom phipinfoio]$ phenv compile_app -i

Command arguments

Argument Description
-h
--help
Display a list and description of arguments.
-s <FILE>
--single-pyfile <FILE>
--single_pyfile <FILE>
Compile a single python file.
-i
--install-app
--install_app
Install the app after it is compiled.
-t
--create-tarball
--create_tarball
Create the app's TAR file and exit. Do not compile or install the app.
-c
--compile-app
--compile_app
Compile the app then exit.
-a <APP_DIR>
--app-dir <APP_DIR>
--app_dir <APP_DIR>
Specify the app's directory.
-d
--exclude-flake
--exclude_flake
Do not run flake when compiling the app.
-e
--continue-on-error
--continue_on_error
Stop compiling or other actions when an error occurs.
-g <IGNORE_FILE>
--ignore-file <IGNORE_FILE>
--ignore_file <IGNORE_FILE>
When compiling the app, ignore the files specified in the ignore file. The default ignore file is .compile_app.ignore.
--force-validate
--force_validate
Run app validation regardless of whether or not you are installing the app after it compiles.
--app-python-version {3,2.7}
--app_python_version {3,2.7}
Specify the app's Python version.

Support for Python 2.7 is included to support legacy apps. Where possible, your apps should be written in Python 3.

--target-dir TARGET_DIR Directory in which to create the app's TAR file.
-v {0,1,2,3}
--verbosity {0,1,2,3}
Set the verbosity level of the command's output;
  • 0 - minimal output
  • 1 - normal output
  • 2 - verbose output
  • 3 - very verbose output
--no-color Don't colorize the command output.
--skip-checks Skip system checks.

Compilation and installation messages

On compilation, the script creates the ../phipinfo.tgz TAR file. This is an app installable TAR file that also can be installed from the UI. However during app development, it's faster to install from the command line using the compile_app<x> script.

cd'ing into ./
Compiling: ./ipinfoio_connector.py
Compiling: ./ipinfoio_consts.py
Compiling: ./__init__.py
Validating App Json
  Working on: ./ipinfoio.json
    Looks like an app json
  Processing App Json
  Processing actions
    test connectivity
      No further processing coded for this action
    geolocate ip
      Following required data paths not in output list
        action_result.data
        action_result.summary
    lookup ip
      Following required data paths not in output list
        action_result.data
        action_result.summary
Installing app...
  Creating tarball...
  ../phipinfoio.tgz
  Calling installer...
[phantom@phantom phipinfoio]$

Explore the installed app

Now that the app has been installed, you can explore the app on the platform.

App table

  1. Go to the Main menu.
  2. Select Apps and search for ipinfo in the Unconfigured Apps tab.
  3. Expand the supported actions node to reveal the action names and their descriptions:
    • lookup ip returns the resolved host name.
    • geolocate ip queries service for IP location information.
    • test connectivity validates the asset configuration for connectivity using supplied configuration.

In , every app requires an asset configuration to be able to run actions on it.

  1. Create an asset by clicking the Configure New Asset button.
  2. Fill in the asset name as "ipinfo" and the asset description as "ipinfo service" on the Asset Information tab.
  3. Click the Asset Settings tab. The Base URL field is auto-filled with https://ipinfo.io.
  4. Click Save.
  5. Click Test Connectivity. This results in the test connectivity app action getting run. The action will fail with an error that says the action is not yet implemented. This is expected, since the default action handler added by the App Wizard will set the status to 'Fail' with the not yet implemented status string.
  6. Click the Documentation link next to the Version value in the app table. This documentation is auto generated by the platform by processing the app JSON. It lists the app details, the contents of the readme.html file, the asset configuration variables, the action listing and the action details.

Run the app actions

You have now defined a working framework of the app and its actions. It is installed on the platform and the user can run the actions from the Investigation page and write playbooks.

Run both of these actions from the Investigation page and show that they both fail with the error action not yet implemented.

For example, do the following to run the geolocate ip action on an event:

  1. Navigate to the Main menu.
  2. From the Sources drop-down menu, select My Events.
  3. Click an event. Take note of the event ID as you will use this when running the playbook.
  4. Select Analyst view.
  5. Click Action.
  6. Click geolocate ip.
  7. Click ipinfo.
  8. Enter the IP address to locate.
  9. Click Save.
  10. Click Launch.

The failure message displays under recent activity.

Create a playbook

Create a playbook for the two actions. The playbook runs both of these actions, and hardcodes in the IP address.

  1. Navigate to the main menu.
  2. Click Playbooks.
  3. Click +Playbook.
  4. From the Start block, drag and release the green node to get started.
  5. Click Action and select ipinfo io from Available Apps.
    1. Click geolocate ip from Available Actions.
    2. Click ipinfo to configure specific assets.
  6. From the Start block, drag and release the green node.
  7. Click Action and select ipinfo io from Available Apps.
    1. Click lookup ip from Available Actions.
    2. Click ipinfo to configure specific assets.
  8. From the Investigate blocks, drag and release the green node to end.
  9. Click Save.

Run the playbook

  1. Navigate to the main menu.
  2. Click Playbooks.
  3. Click on Playbook Debugger.
  4. For the ID field, use the same event ID from when you ran these actions. See Running the app actions.
  5. Click Test.

This playbook will fail when you run it, since you still need to implement the actions. However, save this for now and come back to it later, since it's a useful tool for quickly checking that all of our app actions are properly implemented.

Adding code to the app

Start editing code in /home/phantom/phipinfoio/ipinfoio_connector.py.

_handle_test_connectivity()

Begin by updating the test connectivity handler to include the following:

  • revise self.save_progress
  • revise ret_val, response
  • uncomment return action_result.get_status()
  • uncomment self.save_progress("Test Connectivity Passed")
  • uncomment return action_result.set_status(phantom.APP_SUCCESS)

The updates look as follows:

def _handle_test_connectivity(self, param):

    # Add an action result object to self (BaseConnector)
    # to represent the action for this param
    action_result = self.add_action_result(ActionResult(dict(param)))

    self.save_progress("Testing by sending request to /region")

    ret_val, response = self._make_rest_call('/region', action_result)

    if (phantom.is_fail(ret_val)):
        # the call to the 3rd party device or service failed, action result
        # should contain all the error details
        # so just return from here
        self.save_progress("Test Connectivity Failed")
        return action_result.get_status() 

    # Return success
    self.save_progress("Test Connectivity Passed")
    return action_result.set_status(phantom.APP_SUCCESS)

Most of this code was already present in the template file. The _make_rest_call(...) provides a wrapper around the Python requests module and it parses the output and catches any errors from the request. There are still some changes to make before test connectivity can successfully pass.

_make_rest_call()

The _make_rest_call(...) has some optional parameters that you can use.

def _make_rest_call(self, endpoint, action_result, method="get", **kwargs):
    # **kwargs can be any additional parameters that requests.request accepts

    config = self.get_config()

    resp_json = None

    try:
        request_func = getattr(requests, method)
    except AttributeError:
        return RetVal(action_result.set_status(phantom.APP_ERROR, "Invalid method: {0}".format(method)), resp_json)

    # Create a URL to connect to
    url = self._base_url + endpoint

    try:
        r = request_func(
                        url,
                        # auth=(username, password),  # basic authentication
                        verify=config.get('verify_server_cert', False),
                        **kwargs)
    except Exception as e:
        return RetVal(action_result.set_status( phantom.APP_ERROR, "Error Connecting to server. Details: {0}".format(str(e))), resp_json)

    return self._process_response(r, action_result)

The method can be any HTTP method, including "get", "post", "patch", and "delete". Optionally, data expects a dictionary, which is used when POSTing data. Optionally, params expects a dictionary that will be converted into a query string. The headers is also a dictionary of headers to be sent.

In a few lines previous to the request_func() call there is the line: url = self._base_url + endpoint. This is how you can pass "/region" to _make_rest_call(...), and not the entire URL.

initialize()

Set _base_url in the initialize() function. The initialize() function is called on every action run before any handlers automatically. Since your app configuration already has a base_url parameter, retrieve that and set it. Here is what initialize()looks like after your changes and with the unneeded code cleaned out.

def initialize(self):
    # Load the state in initialize, use it to store data
    # that needs to be accessed across actions
    self._state = self.load_state()

    # get the asset config
    config = self.get_config()

    # Access values in asset config by the name

    # Required values can be accessed directly
    self._base_url = config['base_url']

    return phantom.APP_SUCCESS

As both /region and /hostname return the the data as a single line of text, the Content Type header is set to 'text/html'), as opposed to a JSON object. This is unusual for APIs, and the default implementation of _make_rest_call(...) will return an error. The reason for this is that in a majority of the cases an HTML response is indicative of either making a request to an invalid endpoint, or if there is an error between and a connected proxy.

_process_html_response()

In the _process_html_response(...) function regardless of the content of the response, it sets the action_result to phantom.APP_ERROR. You can modify this method to work by adding two new lines of code to return early with an APP_SUCCESS any time you receive a 200 response from ipinfo.io, which indicates a valid response from the server.

def _process_html_response(self, response, action_result):

    # An html response, treat it like an error
    status_code = response.status_code

    # Handle valid responses from ipinfo.io
    if (status_code == 200):
        return RetVal(phantom.APP_SUCCESS, response.text)

    try:
        soup = BeautifulSoup(response.text, "html.parser")
        error_text = soup.text
        split_lines = error_text.split('\n')
        split_lines = [x.strip() for x in split_lines if x.strip()]
        error_text = '\n'.join(split_lines)
    except:
        error_text = "Cannot parse error details"

    message = "Status Code: {0}. Data from server:\n{1}\n".format(status_code,
            error_text)

    
    message = message.replace('{', '{{').replace('}', '}}')
    

    return RetVal(action_result.set_status(phantom.APP_ERROR, message), None)

Also, you know that a valid response of this type will consist of only a single line of text without any HTML tags or anything, so you don't need to worry about using a module like BeautifulSoup to parse it, and using 'response.text' is sufficient.

The RetVal class is a way of returning a tuple. Using a class like this makes it easier for errors to be caught when going through a linter.

With these changes, you can now recompile and run test connectivity, which will successfully pass.

  1. In the Asset Settings tab of your existing ipinfo asset. See App Table.
  2. Click Test Connectivity.

Edit the code in both of your action handlers. The changes you made to get test connectivity to pass will help these actions, so you don't need to do as much work this time.

_handle_geolocate_ip()

In the handler for geolocate ip for this action, you want to use the endpoint "<IP_ADDRESS>/geo". The IP address you want to check is a required parameter, so you can get it in a similar manner to the base_url. With the IP address, you construct the endpoint URI, then pass that to _make_rest_call(...):

# Required values can be accessed directly
ip = param['ip']

# make rest call
ret_val, response = self._make_rest_call("/{0}/geo".format(ip), action_result)

One of the most important aspects of actions is their ability to save data for later use. You can add data to the action_result object that will end up being sent to after the action is done executing.

# Add the response into the data section
action_result.add_data(response)

You can also add a dictionary object as a summary to show important information about the action once it's done executing. This allows the app author to add the most important part of the result in the summary for consumption of the user in the playbook and also get it displayed in the UI at a prominent position. You make use of update_summary({}) to get the current summary back and append new values to it. In this case, you add the key coordinates and set it to the latitude and longitude returned from ipinfo.io in the loc field.

# Add a dictionary that is made up of the most important values from data into the summary
summary = action_result.update_summary({})
summary['coordinates'] = response["loc"]

Here is what your method looks like with everything added:

def _handle_geolocate_ip(self, param):

    # Implement the handler here
    # use self.save_progress(...) to send progress messages back to the platform
    self.save_progress("In action handler for: {0}".format(self.get_action_identifier()))

    # Add an action result object to self (BaseConnector) to represent the action for this param
    action_result = self.add_action_result(ActionResult(dict(param)))

    # Access action parameters passed in the 'param' dictionary
    # Required values can be accessed directly
    ip = param['ip']

    # make rest call
    ret_val, response = self._make_rest_call("/{0}/geo".format(ip), action_result)

    if (phantom.is_fail(ret_val)):
        # the call to the 3rd party device or service failed, action result should contain all the error details
        # so just return from here
        return action_result.get_status()

    # Now post process the data,  uncomment code as you deem fit

    # Add the response into the data section
    action_result.add_data(response)

    # Add a dictionary that is made up of the most important values from data into the summary
    summary = action_result.update_summary({})
    summary['coordinates'] = response["loc"]

    # Return success, no need to set the message, only the status
    # BaseConnector will create a textual message based off of the summary dictionary
    return action_result.set_status(phantom.APP_SUCCESS)

_handle_lookup_ip

You can do the same thing for the lookup ip handler, noting that you want to send a request to /hostname as opposed to /geo. The action result only accepts a dictionary or JSON object. With the changes to _process_html_response(...), the data you get back is raw text. You need to construct a dictionary and add it to the action result.

# Add the response into the data section
response_dict = {'host_name': response}
action_result.add_data(response_dict)

The response is simple, so update the summary with that same key or value pair.

# Add a dictionary that is made up of the most important values from data into the summary
summary = action_result.update_summary({})
summary['host_name'] = response

With all that added, your handle for lookup ip is similar to this:

def _handle_lookup_ip(self, param):

    # Implement the handler here
    # use self.save_progress(...) to send progress messages back to the platform
    self.save_progress("In action handler for: {0}".format(self.get_action_identifier()))

    # Add an action result object to self (BaseConnector) to represent the action for this param
    action_result = self.add_action_result(ActionResult(dict(param)))

    # Required values can be accessed directly
    ip = param['ip']

    # make rest call
    ret_val, response = self._make_rest_call("/{0}/hostname".format(ip), action_result)

    if (phantom.is_fail(ret_val)):
        # the call to the 3rd party device or service failed, action result should contain all the error details
        # so just return from here
        return action_result.get_status()

    # Now post process the data,  uncomment code as you deem fit

    # Add the response into the data section
    response_dict = {'host_name': response}
    action_result.add_data(response_dict)

    # Add a dictionary that is made up of the most important values from data into the summary
    summary = action_result.update_summary({})
    summary['host_name'] = response

    # Return success, no need to set the message, only the status
    # BaseConnector will create a textual message based off of the summary dictionary
    return action_result.set_status(phantom.APP_SUCCESS)

Recompile the app before running the playbook again.

Python 3:

[phantom@phantom phipinfoio]$ phenv python /opt/phantom/bin/compile_app.pyc -i

Python 2:

[phantom@phantom phipinfoio]$ phenv python2.7 /opt/phantom/bin/py2/compile_app.pyc -i</pre</div>

===Run the playbook again===
Since you've implemented the actions and recompiled the app, you can run your playbook again. This time, both of the actions pass.

#Click on '''Playbook Debugger'''.
# For the ID field, use the same event ID from when you ran these actions. See [[#Running_the_App_Actions|Running the app actions]].
# Click '''Test'''.

===create_output.pyc===
Now that you've successfully run both of the actions, you can use the <code>create_output.pyc</code> script to automatically populate the output list with values. First, find the app run ID for one of the actions. As an example, use the geolocate ip action to illustrate this. 

#In '''Recent Activity''', click on a success message.
#The App Run ID is in the first column of the display. 

With the app run ID, you have all the information you need to update the output.

====Command for updating output with Python 2.7====
Notice the usage of the <code>phenv python2.7</code> and the <code>/bin/py2</code> path to the script. 
<div class=samplecode>
<pre>
[phantom@appdev phipinfoio]$ phenv python2.7 /opt/phantom/bin/py2/create_output.pyc --username admin --password password --app_run_id 1 --app_json ipinfoio.json
Created backup: ipinfoio.json.bak
Updating action: geolocate ip
Successfully updated action output
[phantom@appdev phipinfoio]$

Command for updating output with Python 3

Notice the usage of the phenv python and the /bin path to the script.

[phantom@phantom phipinfoio]$ phenv python /opt/phantom/bin/create_output.pyc --username admin --password password --app_run_id 1 --app_json ipinfoio.json

You don't need to enter the password using the CLI. You can provide your password by setting the PHANTOM_UI_PASSWORD environment variable, or by leaving the parameter blank, at which point the script will prompt you for the password.

Create a table widget

You don't need to copy anything into the app JSON manually. But, you aren't done yet. On the Investigation page, the table widget for the app is still empty. In order to populate it with data, you need to select which data from the output you want to display.

You can add column_name and column_order to your /home/phantom/phipinfoio/ipinfoio.json file and it will be rendered to the table widget.

The App Wizard will include some of the output fields that it knows about into the widget when it's created, but in most cases those aren't going to be sufficient. In this case, since it's geolocate IP, you want to add information pertinent to that. In the case of this API, all of the fields are okay in the output. Other APIs will return much more information than this so you need to be selective about those you want to display.

After adding to the output in ipinfoio.json, and removing the ones the App Wizard created, this is what we will have left:

...
"output": [
    {
        "data_path": "action_result.data.*.ip",
        "data_type": "string",
        "example_values": [
            "8.8.8.8"
        ],
        "contains": [
            "ip"
        ],
        "column_name": "IP",
        "column_order": 0
    },
    {
        "data_path": "action_result.data.*.country",
        "data_type": "string",
        "example_values": [
            "US"
        ],
        "column_name": "Country",
        "column_order": 1
    },
    {
        "data_path": "action_result.data.*.region",
        "data_type": "string",
        "example_values": [
            "California"
        ],
        "column_name": "Region",
        "column_order": 2
    },
    {
        "data_path": "action_result.data.*.city",
        "data_type": "string",
        "example_values": [
            "Mountain View"
        ],
        "column_name": "City",
        "column_order": 3
    },
    {
        "data_path": "action_result.data.*.loc",
        "data_type": "string",
        "example_values": [
            "37.3860,-122.0840"
        ],
        "column_name": "Location",
        "column_order": 4
    }
    ...
]

Make sure the ipinfoio.json is the only JSON file in the directory. Recompile and reinstall the app. See Compile and install the app.

Now look at the output. After reinstalling the app with your changes to the table widget, you don't need to rerun any actions for the changes to show up. The old actions will have their widgets updated.

If the widget is not displaying, do the following from the Investigation page:

  1. Click Manage Widgets.
  2. Slide the ipinfo io toggle to On.
  3. Click Save Layout.
  4. Click Close.

About contains

You may have noticed that both output and action parameters can have a contains value. By matching contains in the output with parameters, playbooks can easily be configured and run based off the output of a previous action.

You can also use contains in the UI. In your own app, for example, you can see how there is an IP in the output, and both actions take an IP as a primary parameter. Because of this, if you click on an actual IP value, a menu appears with various actions that you can run on an IP, including the two from your own app. By clicking on that, it will take you to an action menu with the relevant parameters already filled out.

The create_output.pyc script can add in common contains fields on its own, like 'ip', as well as different types of hashes, URLs, and email addresses. For your own app, there are many times when it makes sense to add in a contains that only that app will use, for example, a ticket ID on a ticketing platform. In that case when adding in the contains, you can follow the convention of "<app name> ticket id", so that there is no conflict with contains from other apps that might have a "ticket id".

Create a custom view

You can also pass data into your own Django templates, which will be rendered as the output. See Use custom views to render results in your app.

Debugging

You can also run actions with a debugger. Open /home/phantom/phipinfoio/ipinfoio_connector.py and look at the bottom of the file, in the main section.

if __name__ == '__main__':

    import sys
    import pudb
    import argparse

    pudb.set_trace()

    argparser = argparse.ArgumentParser()

    argparser.add_argument('input_test_json', help='Input Test JSON file')
    argparser.add_argument('-u', '--username', help='username', required=False)
    argparser.add_argument('-p', '--password', help='password', required=False)

    args = argparser.parse_args()
    session_id = None

    username = args.username
    password = args.password

    if (username is not None and password is None):

        # User specified a username but not a password, so ask
        import getpass
        password = getpass.getpass("Password: ")

    if (username and password):
        try:
            print ("Accessing the Login page")
            r = requests.get("https://127.0.0.1/login", verify=False)
            csrftoken = r.cookies['csrftoken']

            data = dict()
            data['username'] = username
            data['password'] = password
            data['csrfmiddlewaretoken'] = csrftoken

            headers = dict()
            headers['Cookie'] = 'csrftoken=' + csrftoken
            headers['Referer'] = 'https://127.0.0.1/login'

            print ("Logging into Platform to get the session id")
            r2 = requests.post("https://127.0.0.1/login", verify=False, data=data, headers=headers)
            session_id = r2.cookies['sessionid']
        except Exception as e:
            print ("Unable to get session id from the platfrom. Error: " + str(e))
            exit(1)

    if (len(sys.argv) < 2):
        print "No test json specified as input"
        exit(0)

    with open(sys.argv[1]) as f:
        in_json = f.read()
        in_json = json.loads(in_json)
        print(json.dumps(in_json, indent=4))

        connector = TestAppDebugCodeConnector()
        connector.print_progress_message = True

        if (session_id is not None):
            in_json['user_session_token'] = session_id

        ret_val = connector._handle_action(json.dumps(in_json), None)
        print (json.dumps(json.loads(ret_val), indent=4))

    exit(0)

If you look at the code in detail, you may notice the following things:

  • The first few lines are imports followed by the line pudb.set_trace(). This line of code runs a runtime breakpoint and brings up the PuDB debugger.
  • This code expects as input a JSON file that contains data like the configuration and action parameters, which it tries to load: in_json = json.loads(in_json)
  • The input JSON is then dumped on the console print(json.dumps(in_json, indent=4)) for debugging purposes.
  • Next, the IpinfoIoConnector() class object is created connector = IpinfoIoConnector()
  • Some member variables of this object are set, and the _handle_action member function is called with the JSON as input: ret_val = connector._handle_action(json.dumps(in_json), None)
  • Finally, the return value of the _handle_action is dumped on the console before exiting.
  • The code also allows passing the username and password as optional command line switches.

A few things to note here:

  • Running the ipinfoio_connector.py script under a debugger runs the app in standalone mode. In other words, for the most part, it does not talk with . Any progress messages that the app sends are dumped on the console and not sent to the platform. This type of execution helps in debugging the app code that interacts with the external device, in this case the ipinfo.io servers.
  • The code connector._handle_action(json.dumps(in_json), None), notice the _ (underscore) prefixed to the _handle_action function. The ___handle_action__ function is implemented in the BaseConnector class, which does pre-processing with the input JSON including parameter presence validations. If everything looks good, it then calls the IpinfoIoConnector::handle_action function.
  • All REST calls to require a valid session ID. If the action is performing such REST calls to , pass the username and password during the debugging session. The code uses these values to generate a session ID and places it in the passed test JSON before calling the _handle_action function.

The generated code allows setting the password interactively if not specified on the command line.

Debug Environment

Before you can run an action to debug, a few things need to be setup.

Creating a Test JSON

Make use of the create_tj.pyc script to expedite the process of producing a test JSON file. In order to use this script, enable debug logging:

  1. Navigate to the Main menu.
  2. Select Administration.
  3. From the System Health drop-down menu, select Debugging.
  4. From the Action Daemon Log Level drop-down menu, select Debug.
  5. Click Save.

Then create a test JSON based off the most recent run with the specified action name.

Command for creating a test JSON with Python 2.7

Notice the usage of the phenv python2.7 and the /bin/py2/ path to the script.

[phantom@appdev phipinfoio]$ phenv python2.7 /opt/phantom/bin/py2/create_tj.pyc "geolocate ip"
Matched 2 lines, will pick up the last line
IPC json added to file: '/tmp/ipinfoio-geolocate_ip.json'
[phantom@appdev phipinfoio]$

Command for creating a test JSON with Python 3

Notice the usage of the phenv python and the /bin/ path to the script.

[phantom@appdev phipinfoio]$ phenv python /opt/phantom/bin/create_tj.pyc "geolocate ip"

The generated test JSON is going to have more fields than the ones previously specified, but for the sake of testing, those can be safely ignored.

Test JSONs

In the usual scenario, when an action is run by the user through , the platform gathers the asset and action information and other values and creates a JSON that is fed to the BaseConnector::_handle_action function. This performs various validations and pre processing on this input JSON before handing it down to the AppConnector's (such as IpinfoIoConnector) member functions. See Connector module development for more information.

What the test JSONs do is have this input JSON act as if it came from and the code in the main handler calls the BaseConnector::_handle_function directly.

These test JSONs files need to contain the mandatory keys, in addition to the asset configuration and action parameter values.

In the case of the "geolocate ip" action, this JSON looks as follows:

[phantom@appdev phipinfoio]$ cat /tmp/ipinfoio-geolocate_ip.json
{
    "asset_id": "4",
    "connector_run_id": 2,
    "parameters": [
        {
            "ip": "8.8.8.8",
        }
    ],
    "debug_level": 3,
    "action": "geolocate ip",
    "identifier": "geolocate_ip",
    "config": {
        "base_url": "https://ipinfo.io"
    }
}
[phantom@appdev phipinfoio]$
app_config
This key describes the configuration specific to the app. Since the IpinfoIo app does not have any app specific configuration this key is set to null.
asset_id
This key is supposed to contain the ID of the asset that the action is run on. Its value does not matter for debugging purposes. But to get the correct value, in the web UI go to Apps and select the configured asset for ipinfo io. The URL in the browser is of the format https://phantom_ip_address/apps/app_id/asset/asset_id. Copy the numeric ID that is the last part of the URL and paste it as the value of the asset_id key.
config
This is the asset configuration. In this example file the "base_url" is set to "https://ipinfo.io".
debug_level
If nonzero, then the debug_prints will be logged on the console. They will also be logged in the /var/log/phantom/spawn.log file when the actions are run through . The spawn.log file contains both the spawn daemon logs for Python 2 app actions and spawn3 daemon logs for Python 3 app actions.
identifier
This value is the action identifier that was specified in the app JSON. As described in the documentation the action name is different from an identifier. The name of the action is displayed on the product UI, the identifier is used for internal action execution purposes. The main advantage of keeping these two separate is to be able to make a name change without having to change the app code. In this case, the identifier is geolocate_ip.
parameters
This key contains the list of dictionaries where each dictionary represents a single parameter run. Since this action is expecting a single IP address, it contains one element in the parameters list. This can easily be two IPs, and two different action runs, in which case the list looks like the following:
...
...
    "debug_level": 3,
    "identifier": "geolocate_ip",
    "parameters": [
        {
            "ip": "8.8.8.8"
        },
        {
            "ip": "208.67.222.222"
        }
    ]
...

Environment variables

In order to run your app in a debug environment, you also need to set up some environment variables on the OVA.

PYTHONPATH

Whenever an action is run through , the platform sets up the PYTHONPATH so that the BaseConnector and other required modules are available to the App code. In standalone mode this has to be a manual step.

[phantom@appdev phipinfoio]$ export PYTHONPATH=/opt/phantom/lib/:/opt/phantom/www/
[phantom@appdev phipinfoio]$

REQUESTS_CA_BUNDLE

stores the CA certificate bundle in the /opt/phantom/etc/cacerts.pem file. This path is set in the REQUESTS_CA_BUNDLE environment variable, so if an app uses the Python requests module, it directly picks up this bundle. The action path of the bundle is also available to the app through the Python BaseConnector::get_ca_bundle() API. This is a required and manual step while running in standalone mode.

[phantom@appdev phipinfoio]$ export REQUESTS_CA_BUNDLE=/opt/phantom/etc/cacerts.pem
[phantom@appdev phipinfoio]$

As a convenience, you might want to add these to your .bash_profile, so that they are automatically set every time you login. When changed, it will look like the following code:

$ cat .bash_profile
# .bash_profile

# Get the aliases and functions
if [ -f ~/.bashrc ]; then
    . ~/.bashrc
fi

# User specific environment and startup programs

PATH=$PATH:$HOME/bin

export PATH
export PYTHONPATH=/opt/phantom/lib/:/opt/phantom/www/
export REQUESTS_CA_BUNDLE=/opt/phantom/etc/cacerts.pem

Action implementation

After filling up the test JSON with proper values, expand your terminal window to occupy as much of the screen as possible and run the script as follows.

Command for implementing actions with Python 2.7

Notice the usage of the phenv python2.7 and the /bin path to the script.

[phantom@appdev phipinfoio]$ phenv python2.7 /opt/phantom/bin/ipinfoio_connector.py /tmp/ipinfoio-geolocate_ip.json

Command for implementing actions with Python 3

Notice the usage of the phenv python3 and the /py3 path to the script.

[phantom@appdev phipinfoio]$ phenv python3 /opt/phantom/apps/ipinfoio_[guid]/ipinfoio_connector.py /tmp/ipinfoio-geolocate_ip.json

This brings up the PuDB debugger. Since it's been run for the first time, it may bring up a first run dialog box. Since it's running in a console mode, it doesn't support mouse interactions. There are keyboard shortcuts and a quick help dialog box that can be brought up using the ? key.

PuDB

PUDB's first run dialog can be closed by selecting Ok. Optionally, the next screen shows edit preferences for PuDB. You can play around with the theme. Depending on your terminal preferences, you can use the dark vim. Once the theme is modified, quit PuDB using the q key and restart the debugging session.

Once all the dialog boxes are closed, the main window displays the PuDB view which is divided into multiple focus windows:

  • Code
  • Variables
  • Stack
  • Breakpoint
  • Command line

Shortcuts

The following shortcuts are displayed in PuDB when you press the ? key:

  • V - focus variables
  • S - focus stack
  • B - focus breakpoint list
  • C - focus code
  • Ctrl-x - toggle inline command line focus
  • o - minimize the PuDB window and display the console.
  • q - Quit PuDB, can be used in any window focus except Command line (this will result in a bdb.BdbQuit exception that can be safely ignored)

The following shortcuts are available in the code window:

  • n - step over ("next")
  • s - step into
  • c - continue
  • r/f - finish current function
  • t - run to cursor
  • e - show traceback [post-mortem or in exception state]
  • b - set breakpoint at current line

Once in the code window you can move up and down using the the up/down arrows or j/k keys.

  • Keep pressing the n key until you run the print(json.dumps(in_json, indent=4)) line.
  • Pressing the o key takes you to the console view. After minimizing the PuDB window, you can see the input JSON dumped on the console.
  • Press Enter to go back to the PuDB window.
  • Now that the PuDB window is up, navigate to the first line of the function handle_action(self, param) which is ret_val = phantom.APP_SUCCESS and press the b key to set a breakpoint.
  • Press the c key to continue the execution of the rest of the code. For example, call to the BaseConnector::_handle_action which will validate the input JSON.
    • If the input JSON has some errors in it, then the BaseConnector::_handle_action function doesn't call the IpinfoIoConnector::handle_action and return a JSON in ret_val which gets dumped on the console and exit as specified by the code:
      • ret_val = connector._handle_action(json.dumps(in_json), None)
      • print json.dumps(json.loads(ret_val), indent=4)
  • Once the breakpoint in the handle_action function is reached, step through each line of code until you reach the ret_val = self._handle_geolocate_ip(param) line, press the s key to step into this function.

Most of the functions, API, available to an App Connector class are implemented in the BaseConnector class and are accessed through the self object.

App module dependencies

There are two main methods to manage your Python module dependencies in the case your app may need them.

  • Adding the PIP dependencies in the app's JSON
  • Packaged with the app inside a subdirectory called dependencies.

See Specifying pip dependencies for more information.

When you run app actions, the following directories are added to your PYTHONPATH:

  • /opt/phantom/bin
  • /opt/phantom/lib
  • /opt/phantom/www
  • /opt/phantom/apps/<app_install_directory>
  • /opt/phantom/apps/<app_install_directory>/dependencies
Last modified on 18 September, 2024
Use REST handlers to allow external services to call into Splunk SOAR   Frequently asked questions

This documentation applies to the following versions of Splunk® SOAR (Cloud): current


Was this topic useful?







You must be logged into splunk.com in order to post comments. Log in now.

Please try to keep this discussion focused on the content covered in this documentation topic. If you have a more general question about Splunk functionality or are experiencing a difficulty with Splunk, consider posting a question to Splunkbase Answers.

0 out of 1000 Characters