TickTick MCP has clear verb-based naming (add_task, list_tasks, complete_task) and reasonable descriptions (avg ~80 chars). However, parameter descriptions are sparse or missing entirely. For example, add_task's 'priority' parameter lacks explanation of what each integer value means in user terms. Output schemas are not documented, callers cannot predict response structure. Error handling is minimal; most tools return plain strings without guidance on recovery. The server uses fuzzy string matching (e.g., 'searches for matching project') which is fragile and undocumented. No input validation, no enum constraints, no pagination support despite list_tasks accepting a limit parameter.

• Add structured output schemas to all tools. E.g., add_task should return {task_id, title, project_id, priority, due_date} as JSON, not a formatted string. This enables chaining and field extraction.
• Move parameter constraints into descriptions. E.g., 'priority: integer, one of 0 (none), 1 (low), 3 (medium), 5 (high)' and 'due_date: ISO 8601 date (YYYY-MM-DD) or literal 'today'/'tomorrow''. Use enums in schema where possible.
• Document fuzzy matching behavior explicitly. E.g., 'Searches for tasks by title substring (case-insensitive). If multiple matches exist, returns error with list of candidates. Use get_task_details to disambiguate.'
• Add error recovery guidance. E.g., 'Task not found. Try list_tasks() to see available tasks, or use a more specific title substring.'
• Validate inputs early and return actionable errors. E.g., 'Invalid priority: got 2, must be one of: 0, 1, 3, 5' instead of silent API failure.