@pipeworx/gleif

Global Legal Entity Identifier Foundation MCP — LEI lookup, ISIN/BIC-to-entity mapping, and corporate hierarchy, no auth.

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

Tools

• search_lei(query, country?, status?, page_size?) — fuzzy entity-name search. If country narrows the result to nothing, the filter is dropped and the response says so via country_filter_dropped — GLEIF filters on the registered legal address, which often is not where the entity operates (no Sberbank-named record has a Russian one; they sit in CY/SK/LU/GB).

• get_lei(lei) — full Level 1 record: legal name, addresses, jurisdiction, legal form, and LEI registration status. Accepts a company name as well as a 20-character LEI; a resolved name is echoed back as resolved_from_name.

• get_lei_relationships(lei) — Level 2 direct parent, ultimate parent, direct children (one hop).

• isin_to_lei(isin) — resolve a security's ISIN to the legal entity that issued it (LEI + legal name). A valid-but-unmapped ISIN is reported as mapped: false, reason: "isin_unmapped" — the mapping lags new issuance, so unmapped ≠ nonexistent.

• lei_to_isins(lei) — all ISINs issued by an entity (LEI or company name accepted). Returns the full set — one entity issues many securities.

• bic_to_lei(bic) — resolve a SWIFT/BIC (8 or 11 chars) to the bank's legal entity. 8-char codes are padded to the XXX head-office form (GLEIF's filter matches 11-char BICs only); a branch BIC with no mapping of its own falls back to the head office and says so.

• lei_hierarchy_tree(lei, depth?) — multi-hop ownership: ancestor chain up to the ultimate parent plus a breadth-first subsidiary tree (depth 1–4, default 2, node budget 150/call).

Related MCP server: gleif-mcp-server

Data source

https://api.gleif.org/api/v1/ — public JSON:API, no key required. ISIN and BIC joins use GLEIF's official ISIN-to-LEI and BIC-to-LEI mapping files as served through the same API (filter[isin], filter[bic], /lei-records/{lei}/isins).

Passing an LEI

Every tool that takes an lei argument also takes a company name — the name is resolved against the registry and the match is returned alongside the answer. Pass the name whenever that is what you have.

Never construct an LEI code. A guessed 20-character string that happens to be a real LEI belongs to some other company, and GLEIF returns its record with nothing to indicate the wrong entity was fetched — a request for Wirecard AG once came back as Baader Bank Aktiengesellschaft. Codes are checked against the ISO 17442 (ISO 7064 MOD 97-10) check digits before any request is made, and a code that fails is rejected with a message naming search_lei; that catches roughly 96 of every 97 invented codes, but not one that collides with a live record, which is why the name path exists.

Quick Start

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

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

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/gleif/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/search_lei \
  -H 'Content-Type: application/json' \
  -d '{"query":"Apple Inc"}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/search_lei. 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": {
    "gleif": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-gleif"]
    }
  }
}

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-gleif

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 Gleif data" })

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

More

• Docs and guides

• pipeworx.io

License

MIT