Loading ToolBench
Wikidata MCP — wraps Wikidata API (wikidata.org/w/api.php)
Three tools with clear verb-noun naming (search_entities, get_entity, get_wikidata_facts) and solid descriptions (194 - 250 chars, well above 10-char minimum). All tools have input schemas with required fields and type declarations. However, parameter descriptions lack actionable constraints (no enums, ranges, or format specs), output schemas are undocumented, and error handling is minimal. The get_wikidata_facts description includes example values (Q42, Douglas Adams) which LLMs may reuse literally. No pagination support despite potential for large result sets. Schemas are present but sparse, missing properties like minLength, pattern, or enum constraints that would guide LLM input validation.
Get full Wikidata entity by ID (e.g., "Q42" for Douglas Adams, "Q5" for human, "Q1764" for Budapest). Returns labels, descriptions, aliases, claims/statements (properties and values), and sitelinks.
Structured facts about a Wikidata entity in HUMAN-READABLE form — property names and values resolved to labels, not raw P/Q codes. PREFER OVER get_entity for "what is X's <attribute>", "facts about X", "X's date of birth / capital / population". E.g. Q42 (Douglas Adams) -> {"date of birth":["1952-03-11..."],"occupation":["writer",...],"place of birth":["Cambridge"]}. Pass a Q-id from search_entities.
Search Wikidata entities by label or alias (e.g., "Albert Einstein", "Python programming language", "Tokyo"). Returns entity IDs, labels, descriptions, and aliases. Useful for finding the Wikidata ID of any concept.
Output schemas undocumented. LLMs cannot predict response structure (fields, types, nesting). Callers must infer from implementation or trial-and-error.
Parameter descriptions lack actionable constraints. 'language' has no enum (e.g., 'en|fr|de|ja'); 'limit' has no min/max bounds in schema (only in description text). LLMs cannot parse text constraints reliably.
Example values in descriptions (Q42, Douglas Adams, Marie Curie) risk literal reuse by LLMs. Replace with formal constraints (enum, pattern, format) or remove examples entirely.
No pagination support. search_entities returns up to 50 results; get_wikidata_facts caps at 40 properties × 10 values. Large result sets risk context window exhaustion. Missing limit/offset params and total_count in responses.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 59 | 2026-07-28+ | v2 |
Error handling is minimal. Errors are thrown as generic Error objects with brief messages (e.g., 'Wikidata API error (404)'). No recovery guidance, categorization (retryable vs. fatal), or actionable next steps for LLMs.