Loading ToolBench
MCP server for Apple Mail, Calendar, Contacts, Reminders, Notes, and iCloud Drive.
This server has only one tool (hello_apple) with minimal functionality. While the tool has a description and is properly decorated with FastMCP, it lacks critical definition quality: the schema is empty (no parameters documented), descriptions are adequate but generic, and there is no output schema documentation. The tool is a smoke test that returns a string, which is not a production-grade agentic tool. The codebase references a 'tools' module (imported at line 'from . import tools as _tools') but that module is not provided in the source, so actual tool definitions cannot be evaluated. This creates uncertainty about whether the server has additional tools beyond hello_apple that are hidden from this evaluation.
Return macOS version as a smoke test for the local Apple Ecosystem server.
No input schema visible for hello_apple, tool schema score capped at 0
Tool name 'hello_apple' is generic and does not follow verb_noun convention (get_, list_, create_, etc.). 'hello' is not an action verb, it is a greeting. Better names: 'get_macos_version' or 'get_system_version'.
Output schema is not documented. The tool returns a string (macOS version), but the MCP tool definition does not explicitly declare output type, structure, or example format. LLMs cannot plan downstream use without knowing what fields are returned.
Tools module imported but not provided in source code. The server imports 'from . import tools as _tools' but the tools.py file is not included in the evaluation. This prevents verification of tool definitions, schemas, descriptions, and parameter handling for any additional tools the server may expose.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | <=2025-11-25 | v2 |
Description is too brief (under 100 characters) and lacks actionable context. 'Return macOS version as a smoke test' tells the agent it is a test, not when to use it or what the use case is. Add intent guidance: e.g., 'Get the installed macOS version to verify system compatibility or for diagnostic/reporting purposes.'