Skip to content

Security: socialneuron/mcp-server

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

If you discover a security vulnerability in the Social Neuron MCP Server, please report it responsibly.

Email: security@socialneuron.com

Please include:

  • Description of the vulnerability
  • Steps to reproduce
  • Potential impact
  • Suggested fix (if any)

Response timeline:

  • Acknowledgment within 48 hours
  • Initial assessment within 5 business days
  • Fix timeline communicated within 10 business days

Scope

This policy covers:

  • @socialneuron/mcp-server npm package
  • Social Neuron Edge Functions
  • Social Neuron API endpoints

Out of Scope

  • Third-party dependencies (report to upstream maintainers)
  • Social engineering attacks
  • DoS/DDoS attacks

Supported Versions

Version Supported
1.2.x Yes
< 1.2 No

Credential Safety

This npm package contains no service role keys or admin credentials.

  • The SUPABASE_SERVICE_ROLE_KEY is never hardcoded — it is only read from environment variables at runtime, and only in legacy self-hosted mode.
  • The embedded CLOUD_SUPABASE_URL and CLOUD_SUPABASE_ANON_KEY are intentionally public — they are the same values shipped in the frontend bundle. The anon key JWT decodes to "role": "anon", and all data access is gated by Row Level Security (RLS).
  • API keys are stored in the OS keychain (macOS Keychain / Linux secret-tool) or a chmod 0600 credentials file. They are never committed to source control.
  • The npm pack output is restricted to dist/, README.md, CHANGELOG.md, and LICENSE via both .npmignore and package.json files field.

Security Best Practices

  • Always use API key authentication (not service-role keys)
  • Rotate API keys every 90 days
  • Use minimum required scopes (mcp:read for read-only access)
  • Set daily_credit_cap to prevent runaway costs
  • Keep the package updated to the latest version
  • Set DO_NOT_TRACK=1 to disable telemetry if desired

There aren’t any published security advisories