Developer toolkit for Model Context Protocol servers and clients
Critical failures across all dimensions. Tool definitions are severely underdeveloped: all 6 tools have identical, generic descriptions ('...with configurable options') that provide zero actionable guidance. Input schemas are uniformly incomplete, all declare only a generic 'opts' object with no nested property definitions, type constraints, or parameter descriptions. No output schemas are documented. Tool names lack clarity (e.g., 'generate', 'create', 'validate' are generic verbs without context). The codebase shows no evidence of actual tool implementations, examples reference a non-existent McpToolkit class, and src/core.ts is not provided. This appears to be a stub framework rather than a functional MCP server.
Create operation with configurable options
Export operation with configurable options
Generate operation with configurable options
Get templates operation with configurable options
Preview operation with configurable options
Validate operation with configurable options
All tool descriptions are identical boilerplate ('...with configurable options'). Descriptions must explain WHAT the tool does, WHEN to use it, and what it returns. Current descriptions provide zero guidance for LLM tool selection.
Input schemas are incomplete for all tools. All declare only {'opts': {'type': 'object', 'description': '...'}} with no nested properties, types, or constraints. LLMs cannot infer what fields 'opts' accepts or what types they expect.
No output schemas documented. Tools return unstructured results; LLMs cannot plan downstream calls or extract required fields (e.g., IDs for chaining).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 19 | 2026-07-28+ | v2 |
Tool names are generic and lack context. 'generate', 'create', 'validate' do not indicate what is being generated/created/validated. Names should follow verb_noun pattern (e.g., 'generate_report', 'create_user', 'validate_schema').
No error handling guidance. Tools provide no recovery hints, error categorization, or actionable messages. LLMs cannot determine if errors are retryable or user-fixable.