AST-powered MCP server for Vue/React component analysis — eliminates AI hallucinations on local business components
Server provides 11 well-structured tools with complete JSON Schema definitions and descriptions in Chinese. All tools follow verb_noun naming (search_*, get_*, analyze_*) and have documented input parameters with types. However, output schemas are not explicitly documented in the source, parameter descriptions lack actionable constraints (ranges, enums, format specs), and error handling guidance is minimal. No tool annotations (readOnlyHint, destructiveHint) despite all tools being read-only. Descriptions average ~120 chars, within acceptable range but could be more prescriptive about when/why to use each tool.
分析一个页面文件(.vue/.tsx/.jsx),递归解析其完整组件依赖树,包含每个子组件的文件路径和层级关系。帮助 AI 在修改页面前全面理解其架构。
解析组件文件(.vue、.tsx、.jsx),提取其 Props、Events、Slots、依赖项及其他元数据。
在工作区中查找某个组件的所有引用位置(包括 import 导入、模板标签调用及脚本变量引用)。自动感知 unplugin-vue-components 自动注册的组件。
解析单个 Composable / Hook 文件,提取所有 useXxx 函数的完整签名(参数名、类型、默认值)、返回值列表及 JSDoc 描述,帮助 AI 精确理解业务逻辑 Hook。
解析 Vuex/Pinia store 文件,提取其 state、getters、actions 及依赖项(支持 Option Store 和 Setup Store 语法)。
解析单个 TypeScript 类型文件,提取其中所有 interface、type alias、enum 的完整成员定义(字段名、类型、是否可选、JSDoc 注释),帮助 AI 生成类型准确的代码。
使用 ast-grep 的模式匹配语法在工作区搜索代码结构。用于查找函数调用、指令、特定配置等。
Output schemas not documented. Tools return structured data but LLMs cannot see what fields to expect, forcing them to guess downstream field names and risk extraction errors.
Parameter descriptions lack actionable constraints. 'workspacePath' says 'absolute path' but no validation rules (must exist? must be readable?). 'maxDepth' has no min/max bounds stated in description. LLMs cannot self-validate inputs.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 56 | 2026-07-28+ | v2 |
在指定的本地工作区目录中搜索 Vue/React 组件文件(.vue、.tsx、.jsx)。
在工作区的 composables/ 或 hooks/ 目录中搜索所有 useXxx 风格的 Composable / Hook 函数,返回名称、文件路径及参数/返回值数量预览。
在指定的本地工作区目录中搜索 Vuex 或 Pinia store 文件。支持多种目录布局自动检测。
在工作区的 types/、interfaces/ 等目录中搜索所有 TypeScript 类型定义(interface、type alias、enum),返回名称、种类及文件位置预览。
No tool annotations despite all 11 tools being read-only. Missing readOnlyHint in tool definitions. Agents cannot distinguish safe tools from destructive ones without explicit hints.
Error handling in src/index.ts catches all errors and returns generic 'Error processing request: {message}' with isError=true. No recovery guidance, no error classification (retryable vs fatal), no actionable next steps for LLM.
query_ast tool description is vague ('use ast-grep pattern matching'). Does not explain when to use it vs other search tools, what patterns are valid, or what output structure looks like. LLMs may misuse it.