Skip to content

Security: anasali89ji/AI-SCERN

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
Latest (main)
Older releases

Reporting a Vulnerability

Please do not file a public GitHub issue for security vulnerabilities.

Contact

  • Email: security@aiscern.com
  • PGP: (key available on request)
  • Encrypted reports: Preferred for critical vulnerabilities

Scope

In-scope:

  • Authentication and authorization bypasses
  • SQL injection / RLS policy bypasses (Supabase)
  • Remote code execution
  • File upload vulnerabilities (MIME bypass, polyglot files)
  • CSRF vulnerabilities
  • Exposed secrets or credentials in the repository
  • Data leakage between users (cross-account data access)
  • Rate limiting bypasses that enable abuse

Out-of-scope:

  • Denial of service via resource exhaustion without documented proof of impact
  • Social engineering
  • Physical attacks
  • Vulnerabilities in third-party services (Clerk, Supabase, Cloudflare) — report directly to them

Response SLA

Severity Acknowledgment Initial Assessment Fix Target
Critical 24 hours 48 hours 7 days
High 48 hours 5 days 30 days
Medium 5 days 14 days 90 days
Low 14 days 30 days Next release

Safe Harbor

We treat vulnerability research conducted in good faith as follows:

  • We will not pursue legal action for research within scope
  • We will not restrict your access to the platform for reporting
  • We will acknowledge your contribution in our Hall of Fame (with your permission)

Process

  1. Email security@aiscern.com with: severity, affected component, reproduction steps, potential impact
  2. We acknowledge receipt within the SLA above
  3. We investigate and provide a fix timeline
  4. We notify you when the fix is deployed
  5. Coordinated disclosure after fix (90-day embargo unless agreed otherwise)

Hall of Fame

Responsible reporters are recognized at aiscern.com/security/hall-of-fame.


Security Practices

  • All traffic served over HTTPS/TLS 1.3 with HSTS
  • Content-Security-Policy headers on all routes
  • Row-level security (RLS) on all Supabase tables
  • File uploads: MIME allowlist + magic byte validation, UUID-renamed, stored in private R2
  • Rate limiting: Upstash Redis distributed (not in-memory)
  • Secrets: never committed, all via environment variables
  • Dependencies: audited weekly via npm audit in CI

There aren't any published security advisories