The System Designer MCP Server team and community take security vulnerabilities seriously. We appreciate your efforts to responsibly disclose vulnerabilities.

Please do not open a public issue for security vulnerabilities.

To report a security vulnerability, please send an email to: chevonmdphillip@gmail.com

Your email should include:

• A clear description of the vulnerability
• Steps to reproduce the issue (if applicable)
• Potential impact of the vulnerability
• Any suggested mitigation (if known)

• Response Time: We will acknowledge receipt of your vulnerability report within 48 hours.
• Initial Assessment: We will provide an initial assessment of the report within 7 business days.
• Resolution: We will work to address valid security vulnerabilities in a timely manner and aim to release patches in the next scheduled version or sooner if critical.

The System Designer MCP Server implements several security measures:

• All MCP tool inputs are validated using Zod schemas
• Type safety prevents injection attacks
• Input sanitization for file operations

• No arbitrary file system access
• Restricted to designated output directories
• Safe file path handling

• No external network calls initiated by the server
• Optional CORS configuration for web deployments
• No authentication credentials stored in code

• Stateless operation in Cloudflare Workers environment
• No persistent storage of user data
• Sandboxed execution environment

• Keep dependencies updated: bun update
• Use the latest version of the MCP server
• Review tool outputs before execution

• Configure appropriate CORS policies
• Use environment variables for configuration
• Monitor deployment logs for unusual activity
• Consider rate limiting for public deployments

• Validate all inputs before passing to MCP tools
• Use secure communication channels for remote deployments
• Implement proper authentication in client applications

Security updates will be announced through:

• GitHub Security Advisories
• Release notes with security implications clearly marked
• Critical updates may include additional notifications

We welcome responsible security research and value the security community's efforts to improve the safety of our project. If you discover a potential vulnerability, please follow our reporting process above.

The System Designer MCP Server is designed with the following threat model in mind:

1. Input Injection: Malicious inputs attempting to execute arbitrary code
2. Path Traversal: Attempts to access files outside designated directories
3. Resource Exhaustion: Denial of service through excessive resource consumption
4. Data Exposure: Unauthorized access to system or user data

• LLM to Server: The LLM provides structured data, server validates all inputs
• Server to File System: Controlled access to designated directories only
• Server to Network: No outbound network connections initiated by server
• Client to Server: Secure MCP protocol communication

We regularly review and update third-party dependencies to address known vulnerabilities. The project uses:

• @modelcontextprotocol/sdk: MCP protocol implementation
• zod: Schema validation and type safety
• fs-extra: Safe file system operations
• commander: Command-line interface parsing

All dependencies are regularly audited using automated tools and manual review.

For security-related questions that are not vulnerability reports, please:

• Check existing issues and documentation
• Open a regular GitHub issue with the "security" label
• Contact us at chevonmdphillip@gmail.com for sensitive questions

Thank you for helping keep the System Designer MCP Server and its users safe!