Ted Eu MCP server

@pipeworx/ted-eu

TED (Tenders Electronic Daily) MCP — EU public procurement notices from national, regional and municipal buyers in every member state. ~700k notices a year; 1.3M construction notices on file. Keyless.

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

Tools

— newest first by default. category takes plain English ("construction", "IT services", "medical equipment") and resolves to CPV; cpv takes the code if you have it.

• searchnotices(query?, country?, category?, cpv?, datefrom?, dateto?, openonly?, valuemin?, valuemax?, notice_type?, sort?, limit?, page?)

notice_id / id accepted as aliases.

• getnotice(publicationnumber) — one notice with its full buyer-written description.

— contract-AWARD notices only: who won, for how much. Every winner on the notice is returned (lots and consortium members), with winnercountries and decisiondate where published.

• cpv_lookup(term) — plain-English category → CPV code, or what a CPV code covers.
• tedsearchawards(query?, winner?, buyer?, country?, winnercountry?, category?, cpv?, datefrom?, dateto?, valuemin?, value_max?, sort?, limit?, page?)

— everything a named economic operator has won, newest first, with a summary: awarded value by currency, top buyers, top CPV categories. Name matching is partial ("Siemens" covers every Siemens entity).

• tedsupplierhistory(supplier, country?, category?, cpv?, datefrom?, dateto?, limit?, page?)

— everything a named contracting authority has awarded, with top winners and top categories. Use the authority's own-language name ("Ville de Paris", not "City of Paris").

• tedbuyerprofile(buyer, country?, category?, cpv?, datefrom?, dateto?, limit?, page?)

Auth

None. The public search endpoint answers unauthenticated.

Note that GET /notices/{id} is NOT keyless — it 400s with "Missing Authorization header" — so get_notice fetches through the public search endpoint filtered on publication-number instead. Same data, no key.

What comes back

Each notice is flattened out of TED's raw shape:

The envelope carries total, partial and sortedby. partial: true means TED timed out and returned an incomplete result set** — the count and total are lower bounds, not an answer.

Free text

query searches the full notice via TED's FT operator, not just the title. It escalates:

1. the whole string as an exact phrase,
2. its first three words AND-ed, appearing anywhere,
3. its two most specific words AND-ed.

It stops at the first attempt that finds anything and reports which in matchstrategy. This matters because ~ phrase-matches: the old title-or-description phrase search returned zero** for every ordinary multi-word query real callers sent — "backup disaster recovery business continuity", "data center colocation IaaS PaaS", "GRC risk management compliance governance" were all 0 while the AND-ed forms return real notices.

Write OR between two words to hedge a spelling — fuze OR fuse artillery becomes (fuze OR fuse) AND artillery, because AND-ing both spellings asks for the one notice that contains both.

Shortened. The full README is on GitHub.