https://github.com/gpsandhu23/ai_buddy_guard

Prototype AI agent to test how well AI can help us find and fix security problems
https://github.com/gpsandhu23/ai_buddy_guard

Last synced: 3 days ago
JSON representation

Prototype AI agent to test how well AI can help us find and fix security problems

• Host: GitHub
• URL: https://github.com/gpsandhu23/ai_buddy_guard
• Owner: gpsandhu23
• Created: 2023-08-27T19:04:40.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-10-11T21:48:33.000Z (about 1 year ago)
• Last Synced: 2023-10-12T05:17:52.827Z (about 1 year ago)
• Language: Python
• Size: 52.7 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome_ai_agents - Ai_Buddy_Guard - Prototype AI agent to test how well AI can help us find and fix security problems (Building / Testing)
• awesome_ai_agents - Ai_Buddy_Guard - Prototype AI agent to test how well AI can help us find and fix security problems (Building / Security)

README

        
# AI Buddy Guard 🦾

## Overview

Buddy Guard is a prototype AI agent to test how well AI can help us find and fix security problems.

## Features

- **Secret Scanning**: Finds any leaked crdentials in your code

- **Out of date dependencies**: Finds any vulnerable dependencies in your git repo

- **Open S3 buckets**: Checks for any open AWS S3 buckets

- **Missing MFA for AWS**: Find users that don't have MFA enabled on AWS

- **Extract incident schema**: Extract key insights about a security incident in a parsable schema from natural language text reports

- **Check CVE in KEV list**: Check if a CVE is in the CISA KEV list

- **Threat model**: Generate a basic threat model for a service given a url to documentation

- **Invalidate AWS Key**: Invalite an AWS key if it has been compromised

- **Webpage phishing check**: Check if a URL is a phishing webpage based on content, WHOIS records, TLS cert info and DNS records

- **Scan and generate code fixes**: Scan code with GPT-4 and autogenerate code fixes. This is still very noisy, needs more work

## Installation

### Prerequisites

- **Docker**

- **GitHub Personal Access Token for GitHub-related features**

- **AWS tokens for AWS-related features**

### Steps

```bash

# Clone the repository

git clone https://github.com/gpsandhu23/ai_buddy_guard.git

# Navigate to the directory

cd aibuddyguard

# Build and run the Docker container

docker-compose up web_app