Noirlab Datalab MCP Server

@pipeworx/noirlab-datalab

NSF NOIRLab's Astro Data Lab — the DECam, Mayall and Blanco survey catalogues (Legacy Surveys/DECaLS, DES, DELVE, DECaPS, SMASH, NSC) alongside mirrored all-sky catalogues (Gaia DR3, AllWISE, CatWISE, 2MASS, SDSS, DESI), queryable as database tables. Around 130 schemas and tens of billions of rows.

Part of Pipeworx — an MCP gateway connecting AI agents to 1683+ live data sources.

Tools

• datalab_tables(schema?, table?, limit?) — list the survey data releases, one release's tables, or one table's columns. Call it first: table names are survey-specific and a guessed name is a query error, not an empty result.
• datalab_query(sql, limit?) — run a read-only ADQL SELECT. A TOP n is injected when you omit one, so a bare query cannot scan a billion-row table unbounded.
• datalab_cone_search(table, ra, dec, radius?, columns?, limit?) — sources within a radius of a sky position, nearest first, with exact angular separation in degrees.

Auth

Keyless — but only through the IVOA TAP endpoint. See the first trap below.

Data sources

• https://datalab.noirlab.edu/tap/sync — REQUEST=doQuery&LANG=ADQL&FORMAT=csv&QUERY=…. Anonymous, and the only door this pack uses.

Traps

• /query/query?sql= is NOT anonymous, despite being the documented front door. It is what the datalab Python client uses and what the docs lead with, but without a login it answers 401 The provided security token is invalid — and the widely-quoted X-DL-AuthToken: anonymous.0.0.anon_access answers HTTP 200 with the body Error in query(). Try checking that you are logged in. A 200 carrying a failure is the worst shape a failure can take, and it is why this pack speaks TAP only. Measured 2026-09-17.
• TAP reports every error as HTTP 200. Syntax errors, unknown tables, timeouts: all come back 200 with a VOTable whose single INFO element carries QUERY_STATUS="ERROR". Read as data that is an empty result set. Every response is checked for that element before it is parsed.
• ADQL geometry is not implemented, and neither is Q3C. CONTAINS(POINT('ICRS',ra,dec), CIRCLE('ICRS',…)) parses and is handed to PostgreSQL verbatim, which answers function circle(numeric, numeric, numeric) does not exist. The Q3C functions the backend actually has (q3c_radial_query) are rejected by the ADQL parser one layer up. So both obvious cone-search spellings fail. datalab_cone_search builds the search from trigonometry both layers accept: an indexable RA/Dec box (RA widened by 1/cos(dec), wrapping at 0h) to narrow the scan, then an exact great-circle distance to trim the box corners off and order by true separation.
• Row limits are SELECT TOP n, not LIMIT n. ADQL, not PostgreSQL.
• Zero rows is usually footprint, not absence. The DECam surveys cover the southern sky and the Galactic plane selectively, so a northern target legitimately returns nothing from ls_dr10 or des_dr2. The zero-result note says to re-check against gaia_dr3.gaia_source, which is all-sky.

Quick Start

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

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

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/noirlab-datalab/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, 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"
    }
  }
}

Both URLs reach the same gateway and the same 1683+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

curl -X POST https://gateway.pipeworx.io/v1/tools/datalab_query \
  -H 'Content-Type: application/json' \
  -d '{"sql":"SELECT ra, dec, phot_g_mean_mag FROM gaia_dr3.gaia_source WHERE phot_g_mean_mag < 6","limit":5}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/datalab_query. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Standalone (no gateway account)

This package also runs as a local stdio MCP server — no Pipeworx account, no gateway round-trip:

{
  "mcpServers": {
    "noirlab-datalab": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-noirlab-datalab"]
    }
  }
}

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-noirlab-datalab

It speaks MCP over stdin/stdout and answers initialize/tools/list/tools/call for only this pack's tools — none of the shared meta-tools the gateway connection above adds. Same source, same tools, no ask_pipeworx routing.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Noirlab Datalab data" })

The gateway picks the right tool and fills the arguments automatically.

More

• Docs and guides
• pipeworx.io

License

MIT