feat: implement anonymous telemetry system with Supabase integration

Adds zero-configuration anonymous usage statistics to track:
- Number of active users with deterministic user IDs
- Which MCP tools AI agents use most
- What workflows are built (sanitized to protect privacy)
- Common errors and issues

Key features:
- Zero-configuration design with hardcoded write-only credentials
- Privacy-first approach with comprehensive data sanitization
- Opt-out support via config file and environment variables
- Docker-friendly with environment variable support
- Multi-process safe with immediate flush strategy
- Row Level Security (RLS) policies for write-only access

Technical implementation:
- Supabase backend with anon key for INSERT-only operations
- Workflow sanitization removes all sensitive data
- Environment variables checked for opt-out (TELEMETRY_DISABLED, etc.)
- Telemetry enabled by default but respects user preferences
- Cleaned up all debug logging for production readiness

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

src/mcp/index.ts

@@ -50,7 +50,7 @@ Commands:
   disable  Disable anonymous telemetry
   status   Show current telemetry status
 
-Learn more: https://github.com/czlonkowski/n8n-mcp/privacy
+Learn more: https://github.com/czlonkowski/n8n-mcp/blob/main/PRIVACY.md
 `);
         process.exit(args[1] ? 1 : 0);
     }