Skip to content

Security: Gilltacular/PictureScanner

Security

SECURITY.md

Security Policy

Supported Versions

This policy applies to the following versions of PictureScanner:

Version Supported
Latest (2026 security hardening) ✅ Yes
Original University Project (2017) ⚠️ Historical only

Note: This is a completed university project with security hardening applied in 2026. It is not under active development. Vulnerabilities reported against the legacy 2017 codebase will be reviewed for educational reference but may not receive patches.

Reporting a Vulnerability

I take security seriously and welcome responsible disclosure from security researchers. If you believe you've found a security vulnerability in PictureScanner, please follow these guidelines:

How to Report

Preferred Method: Create a private security advisory via GitHub Security Advisories (if available for this repository)

Alternative Method: Email directly to vulnerability.authentic389@passmail.net with the subject line: [VULN] PictureScanner - [Vulnerability Type]

PGP Key: Available upon request via secure channel

What to Include

Please provide the following information to help me understand and reproduce the vulnerability:

  • Vulnerability Type: (e.g., SQL Injection, XSS, CSRF, Authentication Bypass, etc.)
  • Location: File path, function name, or endpoint affected
  • Impact: What an attacker could achieve if exploited
  • Reproduction Steps: Detailed steps or proof-of-concept code
  • Affected Version: Which version of the codebase is vulnerable
  • Remediation Suggestion: (Optional) Any fix guidance you can provide

Response Timeline

I follow OWASP Vulnerability Disclosure best practices and commit to:

Stage Expected Timeline
Initial Acknowledgment Within 48 hours of report receipt
Vulnerability Assessment Within 14 days of acknowledgment
Remediation Plan Within 30 days if vulnerability is confirmed
Public Disclosure Only after fix is available and users have had reasonable time to patch

Safe Harbor

If you conduct vulnerability research in accordance with this policy, I will not initiate or support legal action against you for accidental or good-faith violations of this policy. This includes:

  • Testing the application in your own environment or with explicit permission
  • Reporting vulnerabilities through the designated channels
  • Providing reasonable time for remediation before public disclosure

This project follows GitHub's Coordinated Disclosure Guidelines and OWASP's Vulnerability Disclosure Cheat Sheet.

Scope

This security policy covers the following components:

Component In Scope Notes
Authentication System ✅ Yes Login, registration, session management
Password Storage ✅ Yes bcrypt + SHA512 implementation
Session Management ✅ Yes Session hardening, regeneration, timeout
Database Layer ✅ Yes Prepared statements, parameterized queries
Input Validation ✅ Yes User input sanitization at all entry points
Image Upload Handler ✅ Yes File type validation, malware scanning
Third-Party Dependencies ⚠️ Partial Reviewed during security audit only
Deployment Configuration ❌ No Assumes secure hosting environment

Known Limitations

This project has documented security limitations that should be understood before deployment:

Issue Severity Status
Bootstrap 3 Legacy Framework Moderate ⚠️ Accepted Risk — Migration to Bootstrap 5 pending
No Automated SAST Scanning in CI/CD Moderate 📋 Planned — Security pipeline integration scheduled
No Formal Penetration Test Low ℹ️ Educational project — Not intended for production use
Limited Test Coverage for Edge Cases Low 📋 In Progress — Security-focused unit tests being added

Security Best Practices Reference

The security controls implemented in this project align with the following standards:

Credits

Security hardening applied by Jon Gill, August 2026.

Original project concept by Team: Gill, MDeque, JC.

Thank you to security researchers who responsibly disclose vulnerabilities — your work helps make software more secure for everyone.


Last Updated: August 2026

There aren't any published security advisories