Critical failures across all tools. Descriptions are minimal (10-20 chars) and lack context for LLM selection. Three tools (pizza-map, solar-system, todo-list) have identical, nonsensical parameter schemas (pizzaTopping) that don't match their purpose. get-calculator has zero input parameters despite being a calculation tool. No output schemas documented. No error handling guidance. No parameter descriptions. Naming is weak: 'get-calculator' is vague (get what calculation?), and widget tools lack action verbs.

• Define complete input schemas for each tool. get-calculator needs parameters (e.g., expression: string, precision: integer). pizza-map, solar-system, todo-list should accept meaningful parameters (e.g., theme, size, filter) instead of pizzaTopping.
• Expand descriptions to 50-200 chars. Example: 'Evaluate a mathematical expression and return the result. Supports +, -, *, /, parentheses. Precision defaults to 2 decimal places.' instead of 'Get the result of a calculation based on the provided input'.
• Add descriptions to all parameters. Example for pizzaTopping: 'The pizza topping to highlight in the map widget (string, max 50 chars)', or remove if not actually used.
• Document output schemas in tool descriptions or via a separate schema field. Example: 'Returns {result: number, expression: string, timestamp: ISO8601}' for calculator.
• Rename tools with clear action verbs: get-calculator → evaluate_expression, pizza-map → show_pizza_map, solar-system → show_solar_system, todo-list → show_todo_list.

Inferred effective spec: 2026-07-28+.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry