Automated Web Vulnerability Detection via Virtual Browser Recording
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual penetration testing and existing automated methods often miss vulnerabilities in large web pages and complex business processes, as they are labor-intensive and may not cover all applications or source code thoroughly, leading to incomplete security validation.
Innovation Solution
An automated security testing method that records and plays back user interactions with a web browser to simulate attacks, using macros to discover vulnerabilities and maintain the state of web pages, with a proxy server intercepting and attacking web requests to identify vulnerabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual penetration testing is used to test web page security, then security validation can be performed with human analysis, but vulnerabilities can be missed due to the large size of web pages and the labor-intensive nature of manual analysis
Solution Approach 1:
The patent creates a virtual copy of the web browser environment that can be programmatically controlled. This virtual browser copies the essential functionality of a real browser, allowing automated scripts to interact with web pages, submit forms, and navigate through multiple pages without requiring manual human intervention, thus achieving both comprehensive coverage and high productivity
Solution Approach 2:
The system enables self-service automation where the testing framework automatically performs vulnerability scanning by programmatically controlling the virtual browser to execute test sequences. The system serves itself by automatically generating test cases, executing them against the target web application, and analyzing results without continuous human involvement
2Extent of automation
If automated crawl and audit techniques are used to discover vulnerabilities, then user traffic is not recorded and no authentication data is available to access Web 2.0 applications
Solution Approach 1:
The patent creates a virtual copy of the web browser environment that can be programmatically controlled. This virtual browser copies the essential functionality of a real browser, allowing automated scripts to interact with web pages, submit forms, and navigate through multiple pages without requiring manual human intervention, thus achieving both comprehensive coverage and high productivity
Solution Approach 2:
The system introduces a proxy server as an intermediary between the virtual browser and the target web application. The proxy server intercepts and modifies HTTP requests, injecting attack payloads into form submissions and navigation requests. This intermediary enables automated authentication by capturing and replaying login sequences, allowing the system to access protected Web 2.0 applications while maintaining full automation
3Ease of manufacture
If transport level attacks are used to test for vulnerabilities, then attacks can be created by enclosing parameters within web requests, but the method may not cover complex business processes and source code thoroughly
Solution Approach 1:
The patent transitions from static transport-level attacks to dynamic application-level attacks. The system uses a virtual browser to dynamically generate requests that follow actual business process flows, adapting the attack sequence based on the application's response. This allows the system to cover complex multi-step business processes while maintaining ease of attack creation through automated scripting
Solution Approach 2:
The testing framework segments the vulnerability assessment into distinct phases: authentication phase, navigation phase, form submission phase, and data access phase. Each phase can be independently configured and executed, allowing thorough coverage of complex business processes while maintaining manageable attack creation through modular test case design
Data Source
AI summary
In some examples, a system receives a response from a web server, the response being responsive to a web request sent to the web server. The system executes a script in the response with a web browser, links a document object model (DOM) method to application code executed during the executing of the script, and determines a vulnerability based on the DOM method linked during the executing of the script.