Satellite Imagery

Find the right satellite image of any place and time: the clearest Sentinel-2 or Landsat scene of a field, a city block or a coastline, Sentinel-1 radar when it is cloudy, or NAIP aerial photos of the US, with preview images, true-colour and band download links, and the licence each needs.

A catalogue search returns every scene whose footprint touches your area, including slivers at the edge of a tile, sorted by whatever the catalogue likes. Satellite Imagery measures how much of your area each scene actually covers and ranks by that first, then by cloud cover, then by date (or by clearness or recency if you ask), so the first result is the one you would have picked by hand.

• find_imagery: an area (bounding box, or a point and radius), a date range, a collection and a cloud limit in; ranked scenes out, each with its date, cloud cover, share of your area covered, preview, true-colour image and STAC item link.

• scene_assets: every band and file of one scene, with band name, wavelength, resolution and URL.

It is explicit about what things cost: Landsat, Sentinel-1 and NAIP files sit in requester-pays AWS buckets, so downloading them is billed to the downloader, and those scenes are flagged. Each answer carries the credit line its data provider asks for.

No account or key needed. Built and maintained by Arhan Canli.

Install

Needs Node.js 20 or newer. No account or key.

Claude Code

claude mcp add satellite-imagery -- npx -y satellite-imagery-mcp

Claude Desktop: download satellite-imagery-mcp-<version>.mcpb from the latest release and open it. The bundle is signed; verify it with gh attestation verify <file> --repo arhancanli/satellite-imagery-mcp.

Any other client (Windsurf, Zed, Cline, Continue and others), in its MCP config file:

{
  "mcpServers": {
    "satellite-imagery": {
      "command": "npx",
      "args": [
        "-y",
        "satellite-imagery-mcp"
      ]
    }
  }
}

Docker

docker build -t satellite-imagery-mcp https://github.com/arhancanli/satellite-imagery-mcp.git && docker run -i --rm satellite-imagery-mcp

Hosted (Streamable HTTP): node src/server.mjs --http serves stateless MCP at POST /mcp (port from PORT, default 3000).

Related MCP server: screenshot-website-fast

Example

An agent calls find_imagery with:

{
  "lat": 25.197,
  "lon": 55.274,
  "start": "2026-08-01",
  "end": "2026-09-26",
  "limit": 3
}

and gets back (recorded from the live server on 2026-09-26):

{
  "collection": "Sentinel-2 surface reflectance (Collection 1, L2A), 10 m, every 5 days",
  "area": [
    55.264072,
    25.188017,
    55.283928,
    25.205983
  ],
  "matched": 14,
  "scenes": [
    {
      "id": "S2B_T40RCN_20260830T065854_L2A",
      "date": "2026-08-30T07:02:38",
      "cloud_cover": 0,
      "area_covered_pct": 100,
      "platform": "sentinel-2b",
      "thumbnail": "https://e84-earth-search-sentinel-data.s3.us-west-2.amazonaws.com/sentinel-2-c1-l2a/40/R/CN/2026/8/S2B_T40RCN_20260830T065854_L2A/L2A_PVI.jpg",
      "true_color": "https://e84-earth-search-sentinel-data.s3.us-west-2.amazonaws.com/sentinel-2-c1-l2a/40/R/CN/2026/8/S2B_T40RCN_20260830T065854_L2A/TCI.tif",
      "item": "https://earth-search.aws.element84.com/v1/collections/sentinel-2-c1-l2a/items/S2B_T40RCN_20260830T065854_L2A"
    },
    {
      "id": "S2B_T40RCN_20260919T065429_L2A",
      "date": "2026-09-19T07:02:39",
      "cloud_cover": 0,
      "area_covered_pct": 100,
      "platform": "sentinel-2b",
      "thumbnail": "https://e84-earth-search-sentinel-data.s3.us-west-2.amazonaws.com/sentinel-2-c1-l2a/40/R/CN/2026/9/S2B_T40RCN_20260919T065429_L2A/L2A_PVI.jpg",
      "true_color": "https://e84-earth-search-sentinel-data.s3.us-west-2.amazonaws.com/sentinel-2-c1-l2a/40/R/CN/2026/9/S2B_T40RCN_20260919T065429_L2A/TCI.tif",
      "item": "https://earth-search.aws.element84.com/v1/collections/sentinel-2-c1-l2a/items/S2B_T40RCN_20260919T065429_L2A"
    },
    {
      "id": "S2B_T40RCN_20260909T070216_L2A",
      "date": "2026-09-09T07:02:38",
      "cloud_cover": 0,
      "area_covered_pct": 100,
      "platform": "sentinel-2b",
      "thumbnail": "https://e84-earth-search-sentinel-data.s3.us-west-2.amazonaws.com/sentinel-2-c1-l2a/40/R/CN/2026/9/S2B_T40RCN_20260909T070216_L2A/L2A_PVI.jpg",
      "true_color": "https://e84-earth-search-sentinel-data.s3.us-west-2.amazonaws.com/sentinel-2-c1-l2a/40/R/CN/2026/9/S2B_T40RCN_20260909T070216_L2A/TCI.tif",
      "item": "https://earth-search.aws.element84.com/v1/collections/sentinel-2-c1-l2a/items/S2B_T40RCN_20260909T070216_L2A"
    }
  ],
  "examined": 14,
  "licence": "Copernicus Sentinel data: free and open use; credit 'Contains modified Copernicus Sentinel data [year]'."
}

Tools

Collections: sentinel-2 (Sentinel-2 Collection 1 L2A surface reflectance, 10 m, about every 5 days), landsat (Landsat 8 and 9 Collection 2 Level 2, 30 m), sentinel-1 (radar, through cloud and at night), naip (US aerial, about 1 m, 2010 to 2022).

How it behaves

• Read-only: no tool changes anything outside this process, and nothing is downloaded: results are links.

• Network: HTTPS only, to the hosts listed in package.json under factory.allowHosts, with a deadline, a size cap and bounded retries; a stalled read is retried. Nothing else is contacted, and nothing is logged except unexpected failures (to stderr, without your inputs).

• The 50 best scenes by the catalogue's own order (clearest for optical collections, newest for radar) are ranked, and the answer says when more matched.

• Coverage is computed in longitude and latitude scaled by the cosine of the latitude, which is accurate for the areas scene selection deals with; areas crossing the antimeridian are not supported.

• Results are compact JSON with a matching output schema.

Benchmark

Measured 2026-09-26 with gpt-5.4-mini, 10 fixed tasks graded by fixed checks (bench/tasks.json, raw results in bench/results/).

Performance

Measured 2026-09-26 from Dubai, home connection against the live upstream, Node 24.19.0 (bench/perf.json, scripts/perf.mjs in the factory).

First call: a fresh server process, including the TLS connection and the upstream's own time. Repeat: the same call again, answered from the in-process cache, so it shows this server's own overhead.

Tool definitions the model reads on every turn (name, description, input schema): 1,731 characters, against 2,776 for planetary-computer-mcp (run with mcp<2 pinned; its default install fails to start). The full tool list, with the output schemas and annotations clients use to validate results, is 2,641 characters (4,812 for the alternative).

Data sources

• Earth Search by Element 84, a public STAC API over the AWS Open Data Registry copies of Sentinel-2, Landsat Collection 2, Sentinel-1 and NAIP.

• Licences: Copernicus Sentinel data are free and open (credit "Contains modified Copernicus Sentinel data [year]"); Landsat data from the USGS and NAIP from the USDA are public domain.

More MCP servers by Arhan Canli

• Actions Check: Checks GitHub Actions workflows: outdated actions, old Node runtimes, retired runners, injection.

• Citation Check: Verifies citations: finds fabricated or mismatched references and retractions, returns clean BibTeX.

• Config Check: Validates config files against their official schemas: tsconfig, compose, workflows, 1,400+ more.

• Contact Check: Validates and formats phone numbers, email addresses and postal addresses for any country.

• Cron Check: Explains cron expressions, lists next run times in any time zone, converts between cron dialects.

• Dockerfile Check: Checks Dockerfiles: build-breaking mistakes, base image tags that exist, digests, platforms, EOL.

• Domain Health: Email and domain checks: SPF lookup limits, DKIM keys, DMARC, DNS records, registration expiry.

• Drug Label: FDA drug label answers with section citations, RxNorm name resolution, recalls and shortages.

• The whole collection, 12 more

License

MIT, Copyright (c) 2026 Arhan Canli.