Introduction

  ___            _____ _                    _
 / _ \          |_   _| |                  | |
/ /_\ \_ __  _ __ | | | |__  _ __ ___  __ _| |_
|  _  | '_ \| '_ \| | | '_ \| '__/ _ \/ _` | __|
| | | | |_) | |_) | | | | | | | |  __/ (_| | |_
\_| |_/ .__/| .__/\_/ |_| |_|_|  \___|\__,_|\__|
      | |   | |
      |_|   |_|

This repo builds appthreat/sast-scan (and quay.io/appthreat/sast-scan), a container image with a number of bundled open-source static analysis security testing (SAST) tools. This is like a Swiss Army knife for DevSecOps engineers.

Features

• No messy configuration and no server required
• Scanning is performed directly in the CI and is extremely quick. Full scan often takes only couple of minutes
• Gorgeous HTML reports that you can proudly share with your colleagues and the security team
• Automatic exit code 1 (build breaker) with critical and high vulnerabilities
• There are a number of small things that will bring smile to any DevOps team

Bundled tools

Programming Language	Tools
ansible	ansible-lint
aws	cfn-lint, cfn_nag
bash	shellcheck
bom	cdxgen
credscan	gitleaks
depscan	dep-scan
golang	gosec, staticcheck
java	cdxgen, gradle, find-sec-bugs, pmd
json	jq, jsondiff, jsonschema
kotlin	detekt
kubernetes	kube-score
nodejs	cdxgen, NodeJsScan, eslint, yarn
puppet	puppet-lint
python	bandit, cdxgen, pipenv
ruby	cyclonedx-ruby
rust	cdxgen, cargo-audit
terraform	tfsec
yaml	yamllint

Bundled languages/runtime

• jq
• Golang 1.12
• Python 3.6
• OpenJDK 11 (jre)
• Ruby 2.5.5
• Rust
• Node.js 10
• Yarnpkg

Some reports get converted into an open-standard called SARIF. Please see the section on Viewing reports for various viewer options for this.

Tools enabled for SARIF conversion

• Bash - shellcheck
• Credscan - gitleaks
• Python - bandit
• Node.js - NodeJsScan
• Java - pmd, find-sec-bugs
• Golang - gosec, staticcheck
• Terraform - tfsec

Usage

sast-scan is ideal for use with CI and also as a pre-commit hook for local development.

Integration with Azure DevOps

Refer to the document

Integration with GitHub action

This tool can be used with GitHub actions using this action. All the supported languages can be used.

This repo self-tests itself with sast-scan! Check the GitHub workflow file of this repo.

- name: Self sast-scan
  uses: AppThreat/sast-scan-action@v1.0.0
  with:
    output: reports
    type: python,bash
- name: Upload scan reports
  uses: actions/upload-artifact@v1.0.0
  with:
    name: sast-scan-reports
    path: reports

Integration with Google CloudBuild

Use this custom builder to add sast-scan as a build step.

The full steps are reproduced below.

1. Add the custom builder to your project

git clone https://github.com/CloudBuildr/google-custom-builders.git
cd google-custom-builders/sast-scan
gcloud builds submit --config cloudbuild.yaml .

2. Use it in cloudbuild.yaml

steps:
  - name: "gcr.io/$PROJECT_ID/sast-scan"
    args: ["--type", "python"]

Integration with CircleCI

Refer to the document

Custom integration

SARIF reports produced by sast-scan can be integrated with other compatible tools. It can also be easily imported into databases such as BigQuery for visualization purposes. Refer to integration document for detailed explanation on the SARIF format.

Scanning projects locally

Scan python project

docker run --rm -e "WORKSPACE=${PWD}" -v $PWD:/app appthreat/sast-scan scan --src /app --type python

Scan multiple projects

docker run --rm -e "WORKSPACE=${PWD}" -v $PWD:/app appthreat/sast-scan scan --src /app --type credscan,nodejs,python,yaml --out_dir /app/reports

Scan java project

For java and jvm language based projects, it is important to compile the projects before invoking sast-scan in the dev and CI workflow.

docker run --rm -e "WORKSPACE=${PWD}" -v ~/.m2:/.m2 -v <source path>:/app appthreat/sast-scan scan --src /app --type java

# For gradle project
docker run --rm -e "WORKSPACE=${PWD}" -v ~/.gradle:/.gradle -v <source path>:/app appthreat/sast-scan scan --src /app --type java

Automatic project detection

Feel free to skip --type to enable auto-detection. Or pass comma-separated values if the project has multiple types.

Invoking built-in tools

Bandit

docker run --rm -v <source path>:/app appthreat/sast-scan bandit -r /app

Viewing reports

Reports would be produced in the directory specified for --out_dir. In the above examples, it is set to reports which will be a directory under the source code root directory.

Some of the reports would be converted to a standard called SARIF. Such reports would end with the extension .sarif. To open and view the sarif files require a viewer such as:

• Online viewer - http://sarifviewer.azurewebsites.net/
• VS Code extension - https://marketplace.visualstudio.com/items?itemName=MS-SarifVSCode.sarif-viewer
• Visual Studio extension - https://marketplace.visualstudio.com/items?itemName=WDGIS.MicrosoftSarifViewer
• Azure DevOps extension - https://marketplace.visualstudio.com/items?itemName=sariftools.sarif-viewer-build-tab

Example reports:

Online viewer can be used to manually upload the .sarif files as shown.

Azure DevOps SARIF plugin can be integrated to show the analysis integrated with the build run as shown.

Alternatives

GitLab SAST uses numerous single purpose analyzers and Go based converters to produce a custom json format. This model has the downside of increasing build times since multiple container images should get downloaded and hence is not suitable for CI environments such as Azure Pipelines, CodeBuild and Google CloudBuild. Plus the license used by GitLab is not opensource even though the analyzers merely wrap existing oss tools!

MIR SWAMP is a free online service for running both oss and commercial static analysis for a number of languages simillar to sast-scan. There is a free SWAMP-in-a-box offering but the setup is a bit cumbersome. They use a xml format called SCARF with a number of perl based converters. SARIF, in contrast, is json based and is much easier to work with for integration and UI purposes. By adopting python, sast-scan is a bit easy to work with for customisation.