Ena MCP server

@pipeworx/ena

The European Nucleotide Archive (EMBL-EBI) — the public record of raw sequencing runs, assemblies and annotated sequence, searchable by organism, study, platform, collection country or date, with FASTQ/BAM download URLs and checksums for every run.

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

Tools

result type with ENA's own query grammar (taxeq(2697049), taxtree(9606) AND library_strategy="WGS", country="Kenya"). Answers "what public sequencing data exists for this organism / disease / place".

• ena_search(result?, query?, fields?, sortFields?, limit?) — search any ENA

a study, run, sample or experiment accession, with direct download URLs, byte sizes and MD5s. Answers "where do I get the reads".

• ena_filereport(accession, result?, fields?, limit?) — every data file behind

result type with its live record count, and for one named type, every field it accepts. Call this before composing a query.

• enaresultstypes(result?) — the schema-discovery step: every searchable

Auth

Keyless. No registration step.

Data sources

• — typed search over result types.
• — per-accession file listing.
• — result types + live counts.
• — fields per result type.

Things that cost time to rediscover (measured 2026-09-17)

reads as a single malformed string rather than as an error.

• format=json is not the default. Without it you get TSV, which downstream

read_run that is tens of millions of rows. This pack never sends 0.

• limit=0 means UNLIMITED, the opposite of the usual convention. On

valid-looking, useless answer. The pack sends a default field set per result type.

• Numeric columns are strings on the wire ("read_count":"4547927").
• A search with no fields returns the accession column alone — a

{"message": ...} body naming the accepted regexes. We raise that message; returning an empty list would read as "this study has no runs".

• filereport rejects a mismatched accession with HTTP 200 and a

Quick Start

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

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

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/ena/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"
    }
  }
}

Both URLs reach the same gateway and the same 1679+ 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

Shortened. The full README is on GitHub.