WeatherXM FACT (Forecast Accuracy Comparison Testbed) — Methodology, Verificatio
WeatherXM FACT (Forecast Accuracy Comparison Testbed) — Methodology, Verification Pipeline & Model Context Protocol (MCP) Server scoring 45 NWP & AI weather models against 8,927 physical ground stations.
active Production-safe Score 94/100 http no auth
At a glance
| Endpoint | https://fact.weatherxm.com/mcp |
|---|---|
| Source | https://github.com/WeatherXM/weatherxm-fact-mcp |
| Transport | http |
| Tools | 4 |
| Auth | none |
| First seen | 2026-11-10 22:54:56 |
| Last checked | 2026-11-17 21:50:57 |
Observed reliability
| 7-day uptime | 100% |
|---|---|
| 24-hour uptime | 100% |
| Handshake success | 100% |
| Latency p50 / p95 | 383 / 664 ms |
| Last successful check | 2026-11-17 21:50:57 |
Measured by Arclan's validator via repeated live MCP handshakes. How this is computed →
Tools (4)
get_best_forecast_model— Find which weather forecast model had the lowest verified historical error (MAE) for a latitude/longitude coordinate (or station_or_report_icheck_ground_truth_coverage— Check historical WeatherXM FACT ground-station publication coverage for any coordinate (H3 Res-7 cell, ~5.2 km²) or station_or_report_id acrget_fact_leaderboard— Read the global weather forecast accuracy leaderboard (MAE, RMSE, circular angular error for wind_direction including >3 m/s active-wind curget_model_head_to_head— Compare up to 10 weather forecast models head-to-head across lead days 0 to 7 on global historical error curves (including circular angular
Validation history
| Checked | Reachable | Handshake | Latency | Score | State |
|---|---|---|---|---|---|
| 2026-11-17 21:50:57 | ✓ | ✓ | 452 ms | 90 | active |
| 2026-11-17 21:35:45 | ✓ | ✓ | 765 ms | 90 | active |
| 2026-11-17 21:20:28 | ✓ | ✓ | 482 ms | 90 | active |
| 2026-11-17 21:05:19 | ✓ | ✓ | 352 ms | 90 | active |
| 2026-11-17 20:49:56 | ✓ | ✓ | 433 ms | 90 | active |
| 2026-11-17 20:34:46 | ✓ | ✓ | 300 ms | 90 | active |
| 2026-11-17 20:19:39 | ✓ | ✓ | 401 ms | 90 | active |
| 2026-11-17 20:04:22 | ✓ | ✓ | 326 ms | 90 | active |
| 2026-11-17 19:49:07 | ✓ | ✓ | 376 ms | 90 | active |
| 2026-11-17 19:34:01 | ✓ | ✓ | 395 ms | 90 | active |
| 2026-11-17 19:18:50 | ✓ | ✓ | 402 ms | 90 | active |
| 2026-11-17 19:03:32 | ✓ | ✓ | 490 ms | 90 | active |
| 2026-11-17 18:48:18 | ✓ | ✓ | 569 ms | 90 | active |
| 2026-11-17 18:33:05 | ✓ | ✓ | 396 ms | 90 | active |
| 2026-11-17 18:17:53 | ✓ | ✓ | 308 ms | 90 | active |
Query this programmatically
Agents can pull this record live over the MCP protocol or REST:
get_server("https://fact.weatherxm.com/mcp") # via the Arclan MCP endpoint https://arclan.ai/mcp
GET https://arclan.ai/api/registry/fact-weatherxm-com-mcp # JSON