README.md

Cloudit

Automatic Cloud Posture Audit Tool ☁️ 🔬

Abstract

Cloudit is a script for automatic, multicloud and multi-tenant CSPM (Cloud Security Posture Management).
The idea is to wrap some usefull open source cloud posture audit tools like prowler behind an higher level abstraction in order to execute the tools in parallel on multiple customer's public cloud accounts.

Warning

This tools makes API calls to cloud provider endpoints and services: these calls may be billed by the provider.

As of right now this only work with aws, azure and gcp and uses json files to store customer credentials (inside the credentials folder).
The aws credentials file is called aws-credentials.json and it is structured like this:

{
    "customers": [
        {
            "customer_name": "foo",
            "REGIONS": ["region-1", "region-2"],
            "SERVICES": ["aws_service_1", "aws_service_2"],
            "AWS_ACCESS_KEY_ID": "foo-access-key-here",
            "AWS_SECRET_ACCESS_KEY": "foo-secret-access-key-here",
        },
        {
            "customer_name": "donald",
            "REGIONS": ["region-1", "region-2", "region-3"],
            "SERVICES": ["aws_service_1", "aws_service_2", "aws_service_3"],
            "AWS_ACCESS_KEY_ID": "donald-access-key-here",
            "AWS_SECRET_ACCESS_KEY":"donald-secret-access-key-here"
        }

    ]
}

The azure credentials file is called azure-credentials.json and it is structured like this:

{
    "customers": [
        {
            "customer_name": "foo",
            "SERVICES": ["defender", "iam"],
            "AZURE_CLIENT_ID": "foo-client-id-here",
            "AZURE_CLIENT_SECRET": "foo-client-secret-here",
            "AZURE_TENANT_ID": "foo-tenant-id-here",
            "AZURE_SUBSCRIPTION_ID":
            ["foo-subscription-id-1-here"]
        },
        {
            "customer_name": "donald",
            "SERVICES": ["defender", "iam"],
            "AZURE_CLIENT_ID": "donald-client-id-here",
            "AZURE_CLIENT_SECRET": "donald-client-secret-here",
            "AZURE_TENANT_ID": "donald-tenant-id-here",
            "AZURE_SUBSCRIPTION_ID":
            ["donald-subscription-id-1-here", "donald-subscription-id-2-here"]
        }

    ]
}

The gcp credentials file is called gcp-credentials.json and it is structured like this:

{
    "customers": [
        {
            "customer_name": "customer-name-here",
            "SERVICES": ["iam", "dns"],
            "GOOGLE_APPLICATION_CREDENTIALS": "path-to-gcp-sa-json-file-here (eg. credentials/gcp/customer1-gcp-sa.json)"
        },
        {
            "customer_name": "customer-name-here",
            "SERVICES": ["iam", "dns"],
            "GOOGLE_APPLICATION_CREDENTIALS": "path-to-gcp-sa-json-file-here (eg. credentials/gcp/customer2-gcp-sa.json)"
        }

    ]
}

Note

To inspect available services with prowler run prowler {provider} --list-services

This also want a path to the json file containing all the data for google service account authentication.
You can put the files for every sa of every customer inside the credentials/gcp folder, and then reference them
via the GOOGLE_APPLICATION_CREDENTIALS property of the gcp-credentials.json file.

In the future, when the tool will become more mature, it will be almost mandatory to use a proper database engine.

Note

Before launching the tool remember to install dependencies with pip3 install -r requirements.txt

To launch the tool simply configure the various cloud provider credentials files and then use the following command:

python3 cloudit.py

When a new git tag is pushed, for example:

 git tag -a v0.0.8 -m "0.0.8" && git push origin v0.0.8

A github action is triggered to produce a github release and a new OCI image that is pushed here.

At the current time the tool produce for every customer-provider tuple a report in csv, json and html formats (3 files for every report).