MCP.so
ログイン

DESI MCP Server

@SandyYuan

DESI MCP Server について

mcp server for desi and other surveys

基本情報

カテゴリ

その他

ランタイム

python

トランスポート

stdio

公開者

SandyYuan

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "astro_mcp": {
      "command": "python",
      "args": [
        "server.py"
      ]
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is DESI MCP Server?

An MCP server providing fast, comprehensive access to DESI (Dark Energy Spectroscopic Instrument) survey data through dual access methods: Data Lab SQL queries (default) and SPARCL client (backup). It is designed for AI assistants and researchers to search and retrieve DESI spectroscopic data.

How to use DESI MCP Server?

Install dependencies with pip install mcp sparclclient datalab. Run the server with python server.py. Configure as an MCP client using the JSON snippet provided. Invoke tools via MCP, e.g., search_objects(ra=9.9443, dec=41.7221, object_types=["GALAXY"]) or get_spectrum_by_id(sparcl_id="1270d3c4-9d36-11ee-94ad-525400ad1336", format="full").

Key features of DESI MCP Server

  • Accurate distance sorting for coordinate searches
  • No result limits via Data Lab SQL
  • Fast queries with Q3C spatial indexing
  • Cross-survey access (DESI + BOSS + SDSS) via SPARCL
  • Complete spectral data with wavelength/flux arrays and uncertainties
  • Async support for large datasets (>100k results)

Use cases of DESI MCP Server

  • Search for DESI objects by coordinate (nearest, cone, box)
  • Retrieve full spectrum data by SPARCL UUID for analysis
  • Filter objects by type (galaxy, quasar, star), redshift, and data release
  • Cross‑survey queries including BOSS and SDSS

FAQ from DESI MCP Server

What data does DESI MCP Server cover?

DESI DR1 (∼18+ million spectra), DESI EDR (∼1.8 million spectra), and cross‑surveys BOSS DR16, SDSS DR16 via SPARCL. Sky coverage ∼14,000 sq deg, wavelength range 360–980 nm, spectral resolution R∼2000–5500.

What are the runtime dependencies?

Requires Python packages: mcp, sparclclient, datalab, plus standard library modules (asyncio, json, logging).

Can I query large datasets without limits?

Yes. The default Data Lab SQL method has no result limits; async support is available for datasets larger than 100,000 results.

How do I switch between the two access methods?

By default the server uses Data Lab SQL. To use SPARCL for cross‑survey searches, set use_sparcl_client: true in your tool call.

Does the server sort coordinate search results by distance?

Yes. All coordinate searches are properly sorted by distance for accurate “nearest” results.

コメント

「その他」の他のコンテンツ