Server for researching Italian legislation and administrative court rulings from Normattiva, OpenGA, and GA portal sources.
This server provides a specialized legal research interface with 6 read-only tools accessing Italian legislative and court ruling data. Tool naming is clear and verb-led (search_*, get_*, summarize_*), and all tools have descriptions. However, several definition quality issues prevent a higher score: (1) parameter descriptions lack important constraint information (format, ranges, allowed values), (2) output schemas are not documented in the source, (3) error handling guidance is absent, (4) no tool annotations (readOnlyHint) despite all being read-only, (5) some parameter descriptions are generic or incomplete. The server demonstrates domain expertise and reasonable structure, but falls short of production-grade agentic tool standards. Most tools would benefit from richer parameter validation hints and explicit output schema documentation.
Fetch the full text of an Italian law from Normattiva. Supports fetching entire document or specific articles/sections.
Fetch the full text of a court ruling from the GA portal (Giustizia Amministrativa).
Search Italian administrative court rulings (TAR and Consiglio di Stato) by keywords, court, and date range. Returns metadata from OpenGA.
Search Italian legislation by citation (e.g., 'd.lgs. 196/2003') or keywords. Returns matching laws with metadata.
Full-text search across court rulings with actual ruling content. Slower than search_court_rulings but returns full text snippets.
Aggregate and count court rulings by outcome, court, date period, or other criteria. Returns statistics and trends.
Output schemas not documented. While input schemas are present with type and description for parameters, no tool documents what it returns. LLMs cannot plan downstream calls or extract required fields without knowing return structure.
Missing tool annotations. All 6 tools are read-only (Risk: READ_ONLY noted), but none declare readOnlyHint in their definitions. Agents cannot distinguish safe vs. destructive operations without explicit annotations.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 61 | <=2025-11-25 | v2 |
Parameter constraint documentation is vague. For example, 'limit' parameters have no min/max bounds stated in descriptions (default=20 or 50 mentioned, but no constraints on actual values). Date parameters ('from_date', 'to_date') lack format validation hints in descriptions. Court parameters lack enum documentation of valid codes (source shows mappings like 'tar_lazio', 'cds', but descriptions say only 'Court code (e.g., ...)' without listing options).
No error handling guidance. Tool descriptions do not explain failure modes or recovery steps. What happens when a Normattiva URL is invalid? When OpenGA returns no results? When a court code doesn't exist? Agents cannot self-correct without error classification and suggestions.
Parameter descriptions lack expected return type hints. For example, 'limit' parameter descriptions do not state 'Returns up to X results', only mention default. Agents cannot understand how limiting affects result structure without this clarity.