mcp.json

{
  "$schema": "https://raw.githubusercontent.com/modelcontextprotocol/registry/main/schema/server.json",
  "name": "waveguard",
  "display_name": "WaveGuard Anomaly Detection",
  "description": "Detect anomalies in any data using GPU-accelerated wave physics simulation. Fully stateless — send training + test data in one call. Works on JSON, numbers, text, time series, images.",
  "repository": {
    "url": "https://github.com/gpartin/WaveGuardClient",
    "source": "github"
  },
  "version_detail": {
    "version": "3.1.0",
    "release_date": "2026-02-26"
  },
  "packages": [
    {
      "registry_name": "pypi",
      "name": "WaveGuardClient",
      "version": "3.1.0"
    }
  ],
  "remotes": [
    {
      "transport_type": "http",
      "url": "https://gpartin--waveguard-api-fastapi-app.modal.run/mcp"
    }
  ],
  "tools": [
    {
      "name": "waveguard_scan",
      "description": "Detect anomalies in data using GPU-accelerated wave physics simulation. Send training data (normal examples) and test data (samples to check) in one call. Returns per-sample anomaly scores, confidence levels, and top features explaining each anomaly."
    },
    {
      "name": "waveguard_health",
      "description": "Check WaveGuard API health, GPU availability, version, and engine status."
    }
  ],
  "categories": ["anomaly-detection", "machine-learning", "security", "monitoring"],
  "tags": ["anomaly-detection", "gpu", "wave-physics", "mcp", "python", "stateless"]
}