Server has 3 tools with complete input schemas and descriptions. Naming follows verb_noun convention (read_webpage, search_web, fact_check). Descriptions are present (50-80 chars) but lack WHEN/WHY guidance and dependency hints. All parameters have types and descriptions, meeting baseline requirements. However, output schemas are not documented, LLMs cannot predict response structure. Error handling is minimal (generic API errors with no recovery guidance). No tool annotations (readOnlyHint, idempotentHint). Composition is sound: each tool has one responsibility, but no pagination support for search_web despite potentially large result sets.

• Add output schema documentation to each tool definition. Include example response structure and field descriptions. E.g., 'Returns {content: string, links: [{url, title}], images: [{src, alt}]}'.
• Expand descriptions to 100-150 chars with WHEN/WHY guidance. E.g., 'Extract full webpage content optimized for LLM processing. Use when you need complete article text, metadata, or structured data. Returns markdown, HTML, or plaintext depending on format param.'
• Add readOnlyHint: true and idempotentHint: true to all three tools in the tool definition object returned by ListToolsRequestSchema handler.
• Implement pagination for search_web: add limit (default 10, max 50) and offset parameters. Return total_results count. Document in description: 'Returns up to 50 results per call; use offset for pagination.'
• Enhance error messages with recovery guidance. E.g., 'Jina API returned 401 Unauthorized. Verify JINA_API_KEY is set and valid. If error persists, check https://jina.ai/status for service incidents.'