data.europa.eu — official EU open-data hub (~1.6M datasets)
Server defines 5 tools with basic schemas and minimal descriptions. All tools have input schemas with type definitions, but descriptions are extremely terse (8-27 chars, well below the 50-200 char baseline for LLM-optimized descriptions). Most parameters lack descriptions entirely. No output schemas documented. Error handling is minimal. Tools follow verb_noun naming but descriptions provide almost no context for LLM selection or usage guidance.
List themes/groups.
List publishing organizations.
Single dataset by id.
Search datasets.
List or search tags.
Tool descriptions critically underdeveloped. All 5 tools have descriptions under 30 characters (baseline 50-200 chars). 'Search datasets' and 'List publishing organizations' provide no context for WHEN to use each tool, WHAT it returns, or HOW it differs from similar tools. LLMs cannot reliably select the right tool without richer descriptions.
Parameter descriptions almost entirely missing. Of 11 parameters across 5 tools, only 2 have descriptions ('fq' and 'rows' in search). Parameters like 'query', 'limit', 'start', 'sort', 'id' lack any explanation of expected format, range, or semantics. LLMs cannot infer whether 'limit' means 1-100 or 1-1000 without explicit guidance.
No output schemas documented. Tools return raw API responses (JSON objects with unknown structure). LLMs cannot plan downstream calls or extract required fields without knowing what fields exist, their types, or which are required. This violates the pattern:tool requirement that output schemas be documented.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 45 | 2026-07-28+ | v2 |
Error handling provides no recovery guidance. The deGet() function throws generic errors like 'data.europa.eu: 404' or 'data.europa.eu: 500 <text>'. No indication of whether errors are retryable, what the LLM should do next, or how to correct invalid input. Violates pattern:recovery-guide.
Parameter constraints not enforced or documented. 'rows' accepts 1-1000 but description only says '1-1000 (default 25)', no mention of what happens if LLM passes 5000. 'limit' parameters have no documented bounds. LLMs will pass invalid values without clear feedback on what's acceptable.