Demo

Demo of GuardAI

This demo provides an overview of how the GuardAI tool can be used both in a CI pipeline and locally via the command line.

CI Integration with GitHub Actions

GuardAI can be integrated into your CI pipeline using GitHub Actions/Workflows. In this example, the workflow is configured to run on pull requests targeting the main branch. The action scans the code in the specified directory and generates a report. The workflow then posts the results as a comment on the pull request.

• Workflow File: Demo CI Workflow
• Pull Request Example: #2: Demonstration PR

Pull Request Workflow

1. Run GuardAI Action: The action scans the code in the src directory and outputs the results to a file (guardai_output.txt).
2. Comment on PR: The content of the guardai_output.txt file is automatically posted as a collapsible comment on the pull request, allowing reviewers to easily view the scan results.

Example of how the GuardAI output is commented on a PR:

Local CLI Usage

GuardAI is also designed to be used locally via the command line, allowing developers to scan their code for vulnerabilities before pushing changes to a repository. This ensures that issues can be caught early in the development process.

How to Use GuardAI Locally

1. Run GuardAI from the command line:

   pip install guardai
   
   export OPENAI_API_KEY=<KEY>
   
   guardai --provider openai --directory ./src

2. Review the output: GuardAI will scan the code in the specified directory and output the results directly in your terminal or to a specified output file.

Example of GuardAI CLI usage:

Summary:

For more information, check out the GuardAI repository.