Server has 7 well-named tools with complete input schemas and descriptions. Tool names follow verb_noun pattern (search_analytics, list_sites, get_top_*). All parameters have type definitions and descriptions. However, output schemas are not formally documented in the code, responses are constructed ad-hoc in return statements. Error handling is minimal; no recovery guidance or actionable error messages. Parameter descriptions are good (60-150 chars) but lack constraint details (e.g., no explicit enum for search_type values, no min/max for row_limit). Missing tool annotations (readOnlyHint, destructiveHint). All tools are read-only, which is appropriate, but this should be declared.

• Add formal output schema documentation for each tool. Define return types as TypedDict or JSON Schema (e.g., SearchAnalyticsResult with fields: rows: List[Dict], row_count: int, response_aggregation_type: str). Include in tool docstring or via @mcp.tool() metadata.
• Add tool annotations: @mcp.tool(readOnlyHint=True, idempotentHint=True) to all 7 tools to signal safety to agents.
• Convert free-form string parameters to enums. Change search_type and aggregation_type to Literal types or add JSON Schema enum constraints. Example: search_type: Literal['web', 'image', 'video', 'news', 'discover', 'googleNews'].
• Implement error handling with actionable recovery guidance. Wrap _query() and _service() calls in try-except; return structured errors like {'error': 'Invalid site_url', 'recovery': 'Call list_sites() to find verified sites', 'available_sites': [...]}'.
• Add numeric bounds to parameter descriptions. Update row_limit description to '(1 - 25000, default 1000)' and limit descriptions to '(1 - 1000, default 100)'. Add min/max to JSON Schema if using fastmcp's schema generation.

Inferred effective spec: 2026-07-28+.

• Relies on Dynamic Client Registration (deprecated) - use Client ID Metadata Documents (CIMD)