Skip to content

Security: sinewaveai/agent-security-scanner-mcp

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

We take security seriously. If you discover a security vulnerability in agent-security-scanner-mcp, please report it responsibly.

How to Report

Please do NOT report security vulnerabilities through public GitHub issues.

Instead, please email us at: security@sinewave.ai

Include the following information:

  • Description of the vulnerability
  • Steps to reproduce
  • Potential impact
  • Any suggested fixes (optional)

What to Expect

  • Acknowledgment: Within 48 hours
  • Initial Assessment: Within 7 days
  • Resolution Timeline: Depends on severity, typically 30-90 days

Severity Levels

Severity Description Target Resolution
Critical RCE, data exfiltration 7 days
High Privilege escalation, auth bypass 14 days
Medium Information disclosure 30 days
Low Minor issues 90 days

Supported Versions

Version Supported
2.x.x ✅ Yes
1.x.x ❌ No (upgrade recommended)

Security Best Practices

When using this tool:

  1. Keep Updated: Always use the latest version
  2. Review Prompts: The prompt scanner helps, but review AI instructions manually for sensitive operations
  3. Verify Packages: Use hallucination detection before installing AI-suggested packages
  4. Scan Before Commit: Run security scans before committing code

Past Security Issues

Version Issue Fixed In
< 2.0.0 Command injection in file path handling 2.0.0

Acknowledgments

We thank the security researchers who help keep this project secure. Contributors will be acknowledged here (with permission).

There aren’t any published security advisories