MCP server integrated into movk-nuxt-docs documentation theme, providing tool discovery and filtering capabilities for AI chat and documentation systems
Scoring was not performed
No input schemas visible in source code. Tools declared in useToolCall.ts composable as simple string/function map with no JSON Schema definitions. Per HARD SCORING RULE: tools with no visible input schema must score 0 for schema dimension.
All tool descriptions are under 20 characters and in Chinese only ('列出所有文档页面' = 9 chars, '检索页面' = 4 chars, '列出所有示例' = 6 chars, '获取示例' = 4 chars). Per HARD SCORING RULE: descriptions under 20 chars cannot exceed 0-20 points. Additionally, no English descriptions provided, limiting accessibility to non-Chinese-speaking developers.
Tools are defined in a Vue composable, not as proper MCP tool registrations. The tool definitions are inferred/implicit, not explicitly registered with an MCP server framework. This suggests the MCP integration may be incomplete or the server is not exposing these tools via standard MCP protocol mechanisms.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 29 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 18 | - | v1 |
Parameters 'path' (for get-page) and 'exampleName' (for get-example) have basic descriptions ('页面路径', '示例名称') but no format constraints, enums, length limits, or regex patterns. No information on whether these are required or optional. No validation guidance for LLMs.
No output schemas documented for any tool. Consumers have no way to know what fields get-page returns, what structure list-pages yields, or how to chain results into downstream tool calls. This violates pattern:tool requirement for documented return types.
No error handling guidance. Tools provide no documentation on failure modes, retryability, or how to recover from invalid inputs (e.g., if path does not exist). No actionable error messages or recovery steps.