Three tools with clear verb-noun naming (search, get_judgment, search_by_case) and reasonable descriptions (80-120 chars). All tools have input schemas with typed parameters and descriptions. However, output schemas are not documented in the source code, the tool definitions show inputs but no explicit return type documentation. Parameter descriptions are adequate but lack format constraints (e.g., date format YYYY-MM-DD is mentioned but not enforced via schema pattern). Error handling is present (HTTP errors caught) but lacks recovery guidance for LLMs. No tool annotations (readOnlyHint, etc.) despite all three being read-only operations.

• Add explicit output schema documentation for each tool. Example: 'Returns: {results: [{doc_id: string, case_number: string, judgment_date: string, court: string, summary: string}], total_count: number, page: number}'
• Add readOnlyHint: true to all three tool definitions in the MCP server registration to signal safety and enable LLM optimizations.
• Enhance error messages with recovery guidance. Wrap HTTP errors: 'if (res.statusCode === 404) throw new Error("Document not found. Try search() with case number or keywords instead.")'
• Add JSON Schema pattern constraint for dateFrom/dateTo: '"pattern": "^\\d{4}-\\d{2}-\\d{2}$"' to enforce YYYY-MM-DD format at schema level.
• Document pagination behavior in search description: 'Returns up to 50 results per page. Use page parameter (1-based) to navigate. Response includes total_count for planning.'