Global Forest Watch MCP server

@pipeworx/global-forest-watch

The Global Forest Watch Data API (WRI / UMD) — the catalogue and query surface behind GFW: annual tree cover loss and gain, primary forest extent, GLAD and RADD deforestation alerts, VIIRS/MODIS fire alerts, carbon flux, protected areas, concessions and Aqueduct water risk.

Part of Pipeworx — an MCP gateway connecting AI agents to 1679+ live data sources.

Tools

(382 datasets as of 2026-09). With a query the whole catalogue is walked, capped at 5 upstream pages. Keyless.

• gfw_datasets(query?, limit?, page?) — list or keyword-search the catalogue

resolution, coverage, update frequency, cautions, licence) and version list; with a version, also that version's status, assets and field/band list. Keyless.

• gfw_dataset(dataset, version?) — one dataset's metadata (source, citation,

one dataset version. Needs a caller key.

• gfwquery(dataset, sql, version?, limit?, apiKey) — read-only SELECT over

Auth

Catalogue browsing is keyless and verified so (2026-09-17). The /query endpoint requires an x-api-key and answers 403 "Request is missing valid API key" without one.

No platform key is provisioned. The intended env var when one lands is PLATFORMGFWKEY; sign up at and mint a key with POST /auth/token then POST /auth/apikey on the data API.

platformKeyEnv is deliberately NOT declared while the key is unset, for the same reason as copernicus-cds — see that README's Auth section. Add it in the same change that lands the secret.

Data sources

page[number] / page[size] (JSON:API style, square brackets); the response carries meta.totalitems and meta.totalpages.

• — catalogue. Paging is

metadata. latest is a valid version alias, but it 307-redirects to the concrete version, so the fetch must follow redirects. A bare curl without -L sees a 307 with an empty body, which reads as an outage.

• — dataset metadata.
• — version

columns/bands. Raster datasets report the column name under pixel_meaning rather than name.

• —

The table is always called data.

• — SQL.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "global-forest-watch": {
      "url": "https://gateway.pipeworx.io/global-forest-watch/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/global-forest-watch/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — askpipeworx, discovertools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Shortened. The full README is on GitHub.