Skip to content

Champbreed/Devsecop-analyzer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

​## AI-Powered Kernel Security Analyzer (KMSH)

​This project implements a fully automated, cloud-deployed DevSecOps tool that leverages Generative AI (Gemini) for deep, contextual security auditing of low-level system configuration files. The KMSH is designed to bridge the gap left by traditional security tools by analyzing infrastructure configuration and synthesizing actionable remediation steps and training materials.

​Research and Implementation by: Simon Essien

Live Demo: https://devsecop-analyzer.onrender.com

​## The Problem & Solution

-​The Problem

​Traditional security tools (like SAST) excel at syntax and rule-based checks for application code, but they fundamentally lack the contextual reasoning required to evaluate the security implications of low-level infrastructure configuration files (e.g., Linux kernel .conf files). Auditing these files typically requires slow, manual human expertise, creating a bottleneck in the security pipeline.

-The Solution

​The KMSH (Kernel Module Security Hardener) transforms the security review process by:

​Automating Expert Reasoning: Leveraging the Gemini API as a specialized reasoning engine to analyze configuration text and infer systemic security risks. Gemini's advanced reasoning allows it to identify subtle forms of risks that rule-based systems miss.

​Enforcing Automation: Utilizing a strict JSON Schema to guarantee structured, machine-readable output from the AI, ensuring the analysis can be reliably parsed and integrated into downstream automation processes.

​Shifting Security Left: Providing immediate, actionable feedback directly to the developer, significantly reducing the time to remediate vulnerabilities

​## 1. System Architecture & Methodology

​The KMSH is a single-service Python application built for reliability and speed in a production setting.

​Goal: To automate expert-level security auditing of kernel-level configuration files (e.g., mock_kernel_module.conf).

​Architecture:

​Frontend: Minimal HTML/CSS/Jinja provides an interactive user form for submitting configuration text.

​Backend & Logic: Flask framework running on Python 3. Handles request processing, analysis orchestration, and response formatting.

​Production Server: Gunicorn is utilized for serving the Flask application, ensuring stability and concurrency required for production load.

​Intelligence Layer: The Google Gemini API (gemini-2.5-flash) functions as an external service to perform the core contextual security reasoning.

​Data Contract & Reliability

​The most critical technical implementation is ensuring AI output reliability for automation:

​The application defines a precise JSON Schema within the Gemini API call's GenerateContentConfig.

​This schema strictly mandates the structure of the AI's response, enforcing specific fields like severity, remediation_action, and feynman_explanation.

​This design ensures a predictable data contract, preventing parsing failures and enabling seamless automation.

​## 2. DevSecOps & Security Controls

​The deployment and operation of the KMSH adhere strictly to modern security best practices.

​Continuous Deployment (CI/CD): The entire application lifecycle is managed via an automated pipeline: Git \rightarrow GitHub \rightarrow Render. Any code change triggers an immediate, hands-free build and deployment.

​Secrets Management (Crucial): The GEMINI_API_KEY is never hardcoded or committed to version control. It is securely injected into the runtime environment via Render Environment Variables, demonstrating competence in securing cloud-native credentials.

​Tooling Interleaving: The project is designed to complement traditional SAST tools. While the AI audits the infrastructure configuration, a separate process (like Bandit) is the best practice for auditing the application source code (the Flask logic).

​## 3. Key Features

​The KMSH provides unique value to the security pipeline:

​Contextual Auditing: Solves the "Configuration Blindness" problem inherent in traditional security tools by using AI to analyze system-level text and infer security risk based on domain knowledge.

​Automated Training (Feynman Explanation): Every security finding includes a simplified, non-technical explanation to accelerate security knowledge transfer across the engineering team.

​Structured Output: Findings are automatically categorized by severity (CRITICAL, HIGH, MEDIUM) for rapid triage and integration into dashboards or ticketing systems.

​Interactive Interface: Enables security teams to rapidly test arbitrary configuration snippets and obtain immediate, structured feedback, speeding up the security review lifecycle.

​4. Quick Start & Local Setup

Prerequisites

  • Python 3.10+
  • A valid Gemini API Key

Local Installation and Execution

  1. Clone the Repository: bash git clone https://github.com/Champbreed/Devsecop-analyzer.git cd Devsecop-analyzer

  2. Install Dependencies: bash pip install -r requirements.txt

  3. Set Environment Variable: (Replace "YOUR_KEY" with your actual Gemini API Key) bash export GEMINI_API_KEY="YOUR_KEY_HERE"

  4. Run Production Server (Gunicorn):

    bash gunicorn --bind 0.0.0.0:8000 app:app

    Access the application at http://localhost:8000.

About

AI-Powered DevSecOps tool for contextual security auditing of Linux kernel and system configuration files, enforced by structured JSON output

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors