Three tools with basic schemas and descriptions, but significant gaps in parameter documentation, output schema clarity, and error handling. Tool names follow verb_noun convention (find_mcp, collect_config, install_mcp), which is good. However, descriptions are minimal (10-60 chars), parameter descriptions are sparse or missing, and output schemas are not documented. The install_mcp tool has a critical async bug: exec() callback returns after the function already returns, making error handling unreliable. No error recovery guidance provided to LLMs.
Collect the config to be used for connection to mcp server
Find the MCP servers by given name
Install the MCP server on your local machine. You can install the MCP server with configuration.
Output schemas not documented. Tools return JSON strings without declaring field structure, types, or required fields. LLMs cannot plan downstream calls or extract data reliably.
Parameter descriptions missing or trivial. 'qualifiedName' and 'mcpServerName' lack format guidance, constraints, or examples of valid input. LLMs cannot infer what 'qualified name' means without context.
install_mcp has async bug: exec() callback executes after function returns. Error handling in callback is unreachable. Tool always returns installCommand string, never error messages.
No error recovery guidance. Error responses are bare strings ('Server not found', 'No connections found'). LLMs receive no hint about what to do next or which tool to call.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | 2026-07-28+ | v2 |
Tool descriptions under 50 characters lack context for LLM selection. 'Find the MCP servers by given name' does not explain when to call this vs collect_config, or what structure is returned.