Python 3.8 or Later

MSTICPy requires Python 3.8 or later. If you are running in hosted environment such as Azure Notebooks, Python is already installed. Please ensure that the Python 3.8 (or later) kernel is selected for your notebooks.

If you are running the notebooks locally, you will need to install Python 3.8 or later. The Ananconda distribution is a good starting point since it comes with many of packages required by MSTICPy pre-installed.

Creating a virtual environment


This is an optional step. You will most likely want to do this if you are installing MSTICPy in a local Python installation. If you are using a cloud notebook environment such as Azure ML you will usually not need to create a virtual environment.

MSTICPy has a significant number of dependencies. To avoid conflicts with packages in your existing Python environment you may want to create a Python virtual environment or a conda environment and install the package there.

For standard python use the venv command to do this (there are also several alternatives to venv available).

~$ python -m venv my_env
~$ ./my_env/scripts/activate
(my_env) ~$

For Conda use the conda create command from a conda shell.

(base) c:\users\ian> conda create -n my_env
(base) c:\users\ian> conda activate my_env
(my_env) c:\users\ian>

You should see the name of the environment that you’ve just created and activated in the prompt.


Run the following command to install the base configuation of MSTICPy.

pip install msticpy

or for the latest dev build

pip install git+

Selective Installation - using “extras”

pip supports specification of an additional parameter sequence known as extras. The syntax for this is:

pip install package_name[extra1,extra2,...]

As of version 0.9.0 MSTICPy has its dependencies split into extras. This allows you to install only the packages that you need and avoid the overhead of time and diskspace of dependencies that you do not need.


extras do not affect the which code from MSTICPy is installed - only the external libraries on which certain functions inside MSTICPy need to work.


zshell/MacOS users - you need to escape the first “[“. Otherwise this is interpreted as the start of a file pattern match expression. In other shells such as bash you may find that you need to escape the leading “[” if the extra name matches the pattern of local files in your current directory.

pip install msticpy\[riskiq]

If you are installing a specific version number you should also escape one of the “=” characters

.code:: bash

pip install msticpy[vt]==1.8.2

Extras in MSTICPy

The extras available in MSTICPy are described in the following table:



Install time (increment)

Install time (full)


  • Most functionality (approx 75%)

  • Kqlmagic Jupyter basic



  • Key Vault and keyring storage of settings secrets




  • Azure API data retrieval (subs, resources, Vms, etc.)

  • Azure storage APIs

  • Azure Sentinel APIs (not data query)

  • Also includes “keyvault”




  • Azure Sentinel data queries

  • Kqlmagic Jupyter extended



sentinel (aliases: azsentinel azuresentinel)

  • Combination of core install plus “azure”, “keyvault” and “kql”




  • Timeseries analysis

  • Event clustering

  • Outlier analysis




  • Splunk data queries




  • VirusTotal V3 graph API (default VT lookup is included in base install)




  • RiskIQ Illuminate threat intel provider & pivot functions




  • Includes all of above packages




  • Development tools plus “base”




  • “dev” plus “all”



The installation times here are meant to be indicative of comparative times for installation - they will vary depending on the performance of your computer and network.

The Install time (increment) column shows times relative to the base install (i.e. assuming you’ve already run pip install msticpy). The Install time (full) column shows the time to install the base plus extra. Both columns assume that the following packages are already installed: jupyter, pandas and matplotlib.

If you do not specify an “extra” in your pip install command, the base dependencies for MSTICPy will be installed. This has a lot of functionality such as networking, pivoting, visualization but excludes most dependencies that are specific to a particular data environment like Azure Sentinel or Splunk.

Some of the extras, like “all” and “azsentinel” are combinations of other options collected together as a convenience. You can also specify multiple extras during install, separating them with commas.

pip install msticpy[azure,kql]


when specifying multiple extras, do not leave spaces between the options - just separate with commas.

Missing “extra” exceptions

If you try to use functionality for a component that needs a dependency that you have not installed you will usually get an informative exception message telling you which “extra” option you need to use to enable that feature.

Exception when trying to use a function that is not installed.

To fix this simply run pip install with the “extra” option shown in the exception message:

pip install msticpy[ml]


In some cases you many not get an informative error. We’ve tried to trap all of the cases but if experience a problem with some MSTICPy functionality (especially an ImportError exception, make sure that you have installed the extra that corresponds to the functionality you are trying to use.

Installing in Managed Spark compute in Azure Machine Learning Notebooks

MSTICPy installation for Managed (Automatic) Spark Compute in Azure Machine Learning workspace requires different instructions since library installation is different.


These notebook requires Azure ML Spark Compute. If you are using it for the first time, follow the guidelines mentioned here :Attach and manage a Synapse Spark pool in Azure Machine Learning (preview):

Once you have completed the pre-requisites, you will see AzureML Spark Compute in the dropdown menu for Compute. Select it and run any cell to start Spark Session. Please refer the docs _Managed (Automatic) Spark compute in Azure Machine Learning Notebooks: for more detailed steps along with screenshots. .. _Managed (Automatic) Spark compute in Azure Machine Learning Notebooks:

In order to install any libraries in Spark compute, you need to use a conda file to configure a Spark session. Please save below file as conda.yml , check the Upload conda file checkbox. You can modify the version number as needed. Then, select Browse, and choose the conda file saved earlier with the Spark session configuration you want.

name: msticpy
- defaults
- bokeh
- numpy
- pip:
    - msticpy[azure]>=2.3.1