This project provides an AI-powered cybersecurity toolkit designed for security analysts, red teamers, and blue teamers. It automates common, time-consuming tasks such as traffic analysis, code auditing, and process inspection, leveraging large language models to interpret data and identify potential threats or vulnerabilities.

How It Works

The toolkit integrates with AI models (DeepSeek or Ollama) to perform various security functions. For traffic analysis, it sends packet data to the AI for maliciousness assessment. JS code auditing involves pasting JavaScript snippets for vulnerability detection. Process analysis uses command-line outputs (like tasklist or ps aux) to classify processes as suspicious, security software, or third-party applications. It also converts HTTP requests to Python code, generates regular expressions from examples, detects WebShells and memory-resident malware, and assists in report writing by summarizing vulnerability findings.

Quick Start & Requirements

• Install via pip install -r requirements.txt after cloning the repository.
• Requires Python and configuration of config.py with API keys/URLs for DeepSeek or Ollama.
• Supports Windows, macOS, and Linux.

Highlighted Details

• AI-driven analysis for network traffic, JS code, and process lists.
• Automated generation of Python POCs/EXPs from HTTP requests.
• Advanced WebShell and memory-resident malware detection capabilities.
• Utility for generating regular expressions and aiding in vulnerability report writing.

Maintenance & Community

Developed independently by DeepSeek. Links to Bilibili videos for demonstrations are provided.

Licensing & Compatibility

The repository does not explicitly state a license. Usage is contingent on respecting the author's terms, with a warning against plagiarism and resale.

Limitations & Caveats

The project is described as the "first" AI cybersecurity toolkit from DeepSeek, suggesting it may be in early stages. The README notes that cross-file code auditing is not supported due to context limitations, as files are processed individually.