This server has critical gaps across naming, descriptions, and schemas. Tool names lack action verbs (add-todo, get-todo, remove-todo use hyphens but are acceptable). Descriptions are extremely brief (10-20 chars), well below the 50-200 char baseline for LLM-optimized tools. Parameter descriptions are completely absent, 'text' and 'id' have no context. Schemas exist but are minimal (Zod definitions visible). No output schema documentation. No error handling guidance. No pagination, no structured responses beyond generic text. The get-todo tool returns hardcoded 'All my tasks' with no actual data structure. This is a toy implementation, not production-ready.

• Expand tool descriptions to 80-150 chars. Example: 'add-todo: Add a new task to your todo list. Returns the task ID and confirmation. Use this when the user wants to create a new item to track.'
• Add parameter descriptions. For 'text': 'The task description (1-500 chars, plain text)'. For 'id': 'The numeric task ID to remove (must exist in your list)'.
• Document output schemas. Define what add-todo returns: {task_id: number, text: string, created_at: ISO8601}. Define what get-todo returns: {tasks: [{id, text, created_at}], total_count: number}.
• Implement actual data persistence and return real task lists in get-todo instead of hardcoded strings.
• Add error handling with recovery guidance. Example: 'Task ID 999 not found. Call get-todo to see available task IDs, then retry with a valid ID.'