AWSDefenderGPT is an AI tool designed to identify and rectify cloud misconfigurations by using Open AI GPT models. AWSDefenderGPT can understand complex queries to detect misconfigurations in cloud environments and provide fixes for them.
This tool merges the capabilities of automated deployment and configuration modification using AI, along with cloud SDK tools. As a result, it transforms into an AI-powered cloud manager that helps you ensure the security of the cloud environment by preventing misconfigurations. By centralizing the process, users can effortlessly address misconfigurations and excessively permissive policies in a single stage, simplifying the handling of potential future threats.
Presented at
Developed with ❤️ by INE
- OpenAI
- Docker
- React
- Python 3
- Flask
- SQLite3
- An AWS Account
- AWS Access Key with Administrative Privileges
- OpenAI access key (GPT4 Recommended)
Manually installing AWSDefenderGPT would require you to follow these steps:
(Note: This requires a linux machine with docker installed, with the /bin/bash shell available)
Step 1. Clone the repo
git clone https://github.com/ine-labs/AWSDefenderGPT.gitStep 2. Build the docker image.
cd AWSDefenderGPT
docker build -t awsdefendergpt .Step 3. Launch the container
docker run -p 5000:5000 awsdefendergptRecommended Browser: Google Chrome
Recommended Model: GPT 4
Note: AWSDefenderGPT suggests commands before executing them, and commands are executed when you accept it from the monitor section in the right corner. By default, this feature is activated, and you can deactivate it by clicking the bell icon.
Sherin Stephen, Software Engineer (Cloud), INE sstephen@ine.com
Rishappreet Singh Moonga, Software Engineer (Cloud), INE rmoon@ine.com
Nishant Sharma, Director, Lab Platform, INE nsharma@ine.com (Guidance)
This program is free software: you can redistribute it and/or modify it under the terms of the MIT License.
You should have received a copy of the MIT License along with this program. If not, see https://opensource.org/licenses/MIT.
This tool utilizes AI and may make mistakes. It is not best suited for production-level AWS accounts. Any damages made are the responsibility of the user.