How shivansh4123/malware-slayer works
Project Type and Purpose
• Type: Full-stack web application (monolithic service)
• Purpose: Static malware detection via file uploads, using YARA rules and PE analysis, with a responsive web UI for user interaction.
File Structure
• app.py
• Acts as the Flask web server, routing requests, orchestrating file storage, analysis, and response.
• uploads/
• Persistent storage for incoming user files (e.g., EXE, DOCX, PDF).
• Rule directories (all containing YARA rule files):
• /malware/ – rules for executables & COM files
• /maldocs/ – rules for document malware
• /exploit_kits/ – rules for PDF exploit kits
• /crypto/ – crypto-related signatures
• /packers/ – packer/obfuscator signatures
• Dependencies: Flask (web framework), yara-python (rule engine), pefile (PE analysis).
README for Additional Insights
• Confirms drag-and-drop, multi-file upload UI (HTML5/JS/CSS).
• Custom verdicts based on static rule matches and entropy/section anomalies.
• Runs locally on port 5001.
Main Components
a. Frontend (Browser/UI Layer)
• HTML5 drag & drop area
• JavaScript uploads files via AJAX/HTTP POST
• Displays results (clean vs. malicious) in real time.
b. Web Server (Flask Application)
• HTTP routing: upload endpoint, results endpoint
• Stores files under /uploads
• Invokes analysis modules
c. Storage
• Local filesystem directory /uploads/
• Temp vs. long-term storage.
d. Analysis Engine
• YARA Engine: loads all rules from the five rule directories
• PE Analysis Module: uses pefile to compute entropy, check sections
• Decision logic: merges YARA hits + PE anomalies → verdict
e. Rule Repository
• The five rule directories serve as a “plugin” store of signatures.
f. External Libraries/Services
• yara-python library
• pefile library
• (In future: could integrate antivirus services, databases, or message queues.)
Relationships & Data Flows
• User drags file → Frontend
• Frontend sends HTTP POST → Flask upload endpoint
• Flask writes file → /uploads/
• Flask calls → YARA Engine & PE Analyzer
• Engines read file + rule files → return findings
• Flask applies verdict logic → sends JSON response
• Frontend updates UI with results
Architectural Patterns & Principles
• Monolithic service: single Flask process handling both API and analysis
• Separation of concerns: UI vs. HTTP/API vs. analysis logic vs. storage
• Plugin-style rule loading: dropping a rule file → immediate inclusion
• Synchronous scanning: simple request/response (could evolve to async tasks)
Technologies & Frameworks
• Flask for routing & templating
• yara-python for signature matching
• pefile for Windows PE analysis
• HTML5/JS/CSS for drag-and-drop UI
• Python standard libraries for filesystem I/O