Skip to content

Security: seeincodes/skyfi-mcp

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

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

Email: security@skyfi.com

Do NOT:

  • Open a public GitHub issue for security vulnerabilities
  • Share vulnerability details publicly before a fix is released

Response timeline:

  • Acknowledgment within 48 hours
  • Initial assessment within 5 business days
  • Fix target within 30 days for critical issues

Security Model

Credential Handling

  • API keys are never logged at any layer (request logs, error traces, analytics)
  • In remote mode, credentials arrive per-request in the X-SkyFi-API-Key header and are never stored server-side
  • In local mode, credentials are read from a user-owned config file and held in memory only

Order Confirmation Integrity

  • execute_* tools require a valid quote_id, user_confirmed: true, and idempotency_key
  • This check is enforced server-side at the Zod schema level — it cannot be bypassed by prompt injection

Prompt Injection Mitigation

  • All user-supplied string inputs are treated as data, never instructions
  • String inputs are length-capped (500 chars) and stripped of control characters
  • The server-side confirmation check is the primary defense against injection-driven order placement

Rate Limiting

  • Per-key rate limiting: 100 calls/min, 20 calls/10s burst
  • Daily order cap: 10 orders/day/key
  • Daily spend cap: $10,000/day/key
  • All limits are per-key isolated

Supported Versions

Version Supported
0.1.x Yes

There aren’t any published security advisories