HAIstings

HAIstings is an AI-powered companion designed to help you assess and prioritize Common Vulnerabilities and Exposures (CVEs) within your Kubernetes infrastructure. Drawing inspiration from Agatha Christie's legendary character Arthur Hastings, the crime-solving partner of Hercule Poirot, HAIstings partners with you to ensure robust security measures in your Kubernetes environments.

Overview

HAIstings analyzes vulnerability reports from tools like trivy-operator, generates prioritized reports, and engages in an interactive conversation to refine its recommendations based on your specific context and requirements.

Features

• Vulnerability Prioritization: Automatically prioritizes vulnerabilities based on severity, impact, and context
• Interactive Refinement: Engages in a conversation to gather more context and refine prioritization
• Infrastructure Context: Ingests infrastructure repository information to provide more relevant recommendations
• Persistent Memory: Maintains conversation history across sessions using checkpoints
• Customizable Output: Adjusts recommendations based on user-provided context
• Retrieval-Augmented Generation (RAG): Selectively includes only relevant infrastructure files in the context, reducing overall context size and improving performance

Installation

Prerequisites

• Python 3.12
• Kubernetes cluster with trivy-operator installed
• Properly configured kubeconfig file

Using Poetry

# Clone the repository
git clone https://github.com/stacklok/HAIstings.git
cd HAIstings

# Install dependencies
poetry install

Using pip

pip install haistings

Usage

Basic Usage

Generate a vulnerability report showing the top 25 most critical vulnerabilities:

haistings

Customizing Output

Specify the number of vulnerabilities to show:

haistings --top 30

Providing Context

Provide additional context to improve prioritization:

haistings --notes usercontext.txt

Where usercontext.txt contains information about your infrastructure, such as:

example-service is a very critical service that is internet-facing. We should assign more priority to it.

Flux is critical to our infrastructure, so if it has a vulnerability on anything related to how it processes git requests, then we should assign it very high priority.

Ingesting Infrastructure Repository

Provide your infrastructure repository for additional context:

haistings --infra-repo https://github.com/yourusername/infra-repo --gh-token YOUR_GITHUB_TOKEN

For a specific subdirectory:

haistings --infra-repo https://github.com/yourusername/infra-repo --infra-repo-subdir kubernetes --gh-token YOUR_GITHUB_TOKEN

RAG Configuration

Control the Retrieval-Augmented Generation functionality:

# Disable RAG (use traditional approach)
haistings --use-vectordb false

# Specify maximum number of relevant files per component
haistings --max-relevant-files 10

Persistent Conversations

Use SQLite to persist conversation history:

haistings --checkpoint-saver-driver sqlite

Full Example

haistings --top 30 --notes usercontext.txt --infra-repo https://github.com/yourusername/infra-repo --max-relevant-files 8 --checkpoint-saver-driver sqlite

How It Works

1. Vulnerability Collection: HAIstings connects to your Kubernetes cluster and collects vulnerability reports from trivy-operator.
2. Prioritization: Vulnerabilities are prioritized based on severity (critical vulnerabilities are weighted 10x more than high vulnerabilities).
3. Repository Ingestion: Infrastructure repository files are ingested and stored in a vector database for efficient retrieval.
4. Relevant File Retrieval: Using RAG (Retrieval-Augmented Generation), only the most relevant files for each vulnerability are retrieved based on similarity search.
5. Context Integration: User-provided context and relevant infrastructure files are integrated into the analysis.
6. Report Generation: A prioritized report is generated in a conversational style inspired by Arthur Hastings.
7. Interactive Refinement: HAIstings engages in a conversation to gather more context and refine its recommendations.

Command Line Options

Example Output

# HAIsting's Security Report

## Introduction

Good day! Arthur Hastings at your service. I've meticulously examined the vulnerability reports from your Kubernetes infrastructure and prepared a prioritized assessment of the security concerns that require your immediate attention.

## Summary

After careful analysis, I've identified several critical vulnerabilities that demand prompt remediation:

1. **example-service (internet-facing service)**
   - Critical vulnerabilities: 3
   - High vulnerabilities: 7
   - Most concerning: CVE-2023-1234 (Remote code execution)
   
   This service is particularly concerning due to its internet-facing nature, as mentioned in your notes. I recommend addressing these vulnerabilities with the utmost urgency.

2. **Flux (GitOps controller)**
   - Critical vulnerabilities: 2
   - High vulnerabilities: 5
   - Most concerning: CVE-2023-5678 (Git request processing vulnerability)
   
   As you've noted, Flux is critical to your infrastructure, and this Git request processing vulnerability aligns with your specific concerns.

[Additional entries...]

## Conclusion

I say, these vulnerabilities require prompt attention, particularly the ones affecting your internet-facing services and deployment controllers. I recommend addressing the critical vulnerabilities in example-service and Flux as your top priorities. Should you require any further assistance or have additional context to share, I remain at your service.

Development

Setting Up Development Environment

# Clone the repository
git clone https://github.com/stacklok/HAIstings.git
cd HAIstings

# Install dependencies including development dependencies
poetry install

# Run tests
poetry run pytest

Code Style

This project uses:

• Black for code formatting
• isort for import sorting
• mypy for type checking
• flake8 for linting

# Format code
poetry run black .
poetry run isort .

# Type check
poetry run mypy .

# Lint
poetry run flake8

Future Improvements / TODO

• Custom Vulnerability Scoring: Add support for custom vulnerability scoring based on user-defined criteria beyond just severity.
• Integration with More Scanners: Extend beyond trivy-operator to support other vulnerability scanners.
• Visualization Dashboard: Create a web interface to visualize vulnerability reports and trends over time.
• Automated Remediation Suggestions: Provide specific remediation steps for common vulnerabilities.
• Multi-Cluster Support: Add support for analyzing vulnerabilities across multiple Kubernetes clusters.

License

Apache-2.0