GEIANT Perception MCP

GEIANT Perception exposes geospatial weather, satellite-tile and classification tools with project audit and delegation evidence.

Remote serverstreamable-httpsseTypeScript

What is the GEIANT Perception MCP Server?

GEIANT Perception exposes geospatial weather, satellite-tile and classification tools with project audit and delegation evidence.

Retrieve information for an area

Perception combines weather retrieval, satellite-tile discovery and optional geospatial classification. Its documented weather source uses Open-Meteo data, while tile discovery uses Sentinel-2 STAC metadata and band URLs. H3 cells identify areas for the workflows. A tile metadata result is not automatically a downloaded raster or a completed land-cover classification.

Separate the available backends

The source implements Streamable HTTP and legacy SSE. Weather and tile lookup have different requirements from classification, which uses a configured RunPod GPU endpoint. The documentation describes the Clay embedding tool as planned; this profile does not add it as a verified available tool. The governance functions inspect the project's audit, trust and delegation records.

Check provenance and coverage

Start with a known location and compare the returned weather coverage, image metadata and cell boundary with the requested area. Inspect timestamps and source coverage inside the data itself when choosing observations. A project compliance report contains project evidence, not independent legal certification. Review the underlying audit records and classification configuration before relying on a summarized result in another workflow.

Project documentation

Repository and setup guide

Recorded tools

Check the linked publisher documentation and your connected client's tool list. Names and availability can vary with the server version and account permissions.

ToolWhat it does
perception_weatherRetrieve weather information for a geospatial area.
perception_fetch_tileRetrieve satellite-tile metadata and band URLs.
perception_classifyClassify imagery through a configured model backend.
gns_get_compliance_reportRetrieve project audit and delegation evidence.
gns_get_trust_scoreRead the project trust score.
gns_verify_chainVerify a project audit chain.
gns_roll_epochRoll project audit evidence into an epoch.

How to install the GEIANT Perception MCP Server

{
  "mcpServers": {
    "geiant-perception": {
      "url": "https://packagesmcp-perception-production.up.railway.app/mcp"
    }
  }
}

Configuration

An HTTP MCP client. Classification needs the configured GPU backend; audit and delegation features depend on the deployment configuration.

Frequently asked questions

No. Tile discovery returns STAC metadata and band URLs. Classification uses a separately configured RunPod endpoint.