mentor-mcp-server
导师-mcp-服务器
模型上下文协议 (MCP) 服务器通过 AI 驱动的 Deepseek-Reasoning (R1) 指导功能,为 LLM 代理提供第二意见,包括代码审查、设计评审、写作反馈以及通过 Deepseek API 进行创意集思广益。借助专家的第二意见和切实可行的见解,为您的 LLM 代理奠定成功的基础。
模型上下文协议
模型上下文协议 (MCP) 支持以下之间的通信:
客户端:Claude Desktop、IDE 和其他与 MCP 兼容的客户端
服务器:任务管理和自动化的工具和资源
LLM 代理:利用服务器功能的 AI 模型
Related MCP server: DSH Crew
目录
特征
代码分析
全面的代码审查
错误检测和预防
风格和最佳实践评估
性能优化建议
安全漏洞评估
设计与建筑
UI/UX 设计评论
架构图分析
设计模式建议
可访问性评估
一致性检查
内容增强
写作反馈和改进
语法和风格分析
文档审查
内容清晰度评估
结构性建议
战略规划
功能增强头脑风暴
对方法的第二意见
创新建议
可行性分析
用户价值评估
安装
# Clone the repository
git clone git@github.com:cyanheads/mentor-mcp-server.git
cd mentor-mcp-server
# Install dependencies
npm install
# Build the project
npm run build配置
添加到您的 MCP 客户端设置:
{
"mcpServers": {
"mentor": {
"command": "node",
"args": ["build/index.js"],
"env": {
"DEEPSEEK_API_KEY": "your_api_key",
"DEEPSEEK_MODEL": "deepseek-reasoner",
"DEEPSEEK_MAX_TOKENS": "8192",
"DEEPSEEK_MAX_RETRIES": "3",
"DEEPSEEK_TIMEOUT": "30000"
}
}
}
}环境变量
多变的 | 必需的 | 默认 | 描述 |
DEEPSEEK_API_KEY | 是的 | - | 您的 Deepseek API 密钥 |
DEEPSEEK_模型 | 是的 | deepseek-reasoner | Deepseek 模型名称 |
DEEPSEEK_MAX_TOKENS | 不 | 8192 | 每个请求的最大令牌数 |
DEEPSEEK_MAX_RETRIES | 不 | 3 | 重试次数 |
DEEPSEEK_TIMEOUT | 不 | 30000 | 请求超时(毫秒) |
工具
代码审查
<use_mcp_tool>
<server_name>mentor-mcp-server</server_name>
<tool_name>code_review</tool_name>
<arguments>
{
"file_path": "src/app.ts",
"language": "typescript"
}
</arguments>
</use_mcp_tool>设计评论
<use_mcp_tool>
<server_name>mentor-mcp-server</server_name>
<tool_name>design_critique</tool_name>
<arguments>
{
"design_document": "path/to/design.fig",
"design_type": "web UI"
}
</arguments>
</use_mcp_tool>撰写反馈
<use_mcp_tool>
<server_name>mentor-mcp-server</server_name>
<tool_name>writing_feedback</tool_name>
<arguments>
{
"text": "Documentation content...",
"writing_type": "documentation"
}
</arguments>
</use_mcp_tool>功能增强
<use_mcp_tool>
<server_name>mentor-mcp-server</server_name>
<tool_name>brainstorm_enhancements</tool_name>
<arguments>
{
"concept": "User authentication system"
}
</arguments>
</use_mcp_tool>示例
每个工具的使用和输出的详细示例可以在示例目录中找到:
第二意见示例——身份验证系统需求分析
代码审查示例- 详细的 TypeScript 代码审查,包含安全性和性能见解
设计评论示例- 仪表板设计的综合 UI/UX 反馈
写作反馈示例- 文档改进建议
头脑风暴增强功能示例- 功能构思与实施细节
每个示例都包括请求格式和示例响应,展示了该工具的功能和输出结构。
发展
# Build TypeScript code
npm run build
# Start the server
npm run start
# Development with watch mode
npm run dev
# Clean build artifacts
npm run clean项目结构
src/
├── api/ # API integration modules
├── tools/ # Tool implementations
│ ├── second-opinion/
│ ├── code-review/
│ ├── design-critique/
│ ├── writing-feedback/
│ └── brainstorm-enhancements/
├── types/ # TypeScript type definitions
├── utils/ # Utility functions
├── config.ts # Server configuration
├── index.ts # Entry point
└── server.ts # Main server implementation执照
Apache 许可证 2.0。有关更多信息,请参阅许可证。
Available Tools
5 toolsbrainstorm_enhancementsC
Generates creative ideas for improving a given concept, product, or feature, focusing on innovation, feasibility, and user value.
Inputs: concept (required).
code_reviewC
Provides a code review for a given file or code snippet, focusing on potential bugs, style issues, performance bottlenecks, and security vulnerabilities.
Inputs: file_path, language, code_snippet.
design_critiqueB
Offers a critique of a design document, UI/UX mockup, or architectural diagram, focusing on usability, aesthetics, consistency, accessibility, and potential design flaws.
Inputs: design_document (required), design_type (required).
second_opinionC
Provides a second opinion on a user's request by analyzing it with an LLM and listing critical considerations.
Inputs: user_request (required).
writing_feedbackC
Provides feedback on a piece of writing, such as an essay, article, or technical documentation, focusing on clarity, grammar, style, structure, and overall effectiveness.
Inputs: text (required), writing_type (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Each tool has a clearly distinct purpose targeting different domains: brainstorming, code review, design critique, second opinions, and writing feedback. There is no overlap in functionality, making it easy for an agent to select the appropriate tool without confusion.
All tool names follow a consistent snake_case pattern with a clear verb_noun structure (e.g., brainstorm_enhancements, code_review). This predictability enhances usability and aligns with common MCP conventions.
With 5 tools, the server is well-scoped for its mentoring/feedback purpose. Each tool serves a unique and valuable function, avoiding bloat while covering key areas like code, design, writing, and idea generation.
The toolset covers major feedback domains (code, design, writing, brainstorming) and includes a general second_opinion tool. A minor gap is the lack of a tool for project management or strategic planning feedback, but core mentoring workflows are well-supported.
Maintenance
Related MCP Connectors
A design agent in your coding agent's loop: briefs each screen, reviews UX and visual quality.
DevSpeak is a writing tool for developers, available to your coding agent over MCP. Call `optimize_text` to improve a PR description, commit message, ticket or doc in the language it was written in, or set `correctionsOnly` to fix only spelling, grammar and punctuation. Call `translate_text` to turn rough notes into a technical spec, Jira tickets, an RFC, an API design or a data model for the audience you choose (Junior Dev, Senior Dev, Tech Lead, SRE or Data Engineer). It also translates README files with `translate_markdown`, refines earlier output and lists your history. Requires a DevSpeak API key; a free account includes one key with 50 API requests a month.
AI colleagues that keep your standards, your project and their reasoning between sessions
Memory, reflection and safety for AI agents. Hosted or local, no vector database.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceProvides code generation and completion capabilities using the DeepSeek API, with support for tool chaining and cost optimization.5ISC
- AlicenseNot gradedqualityAmaintenanceEnables dispatching tasks from coding hosts such as Claude Code, Codex, Antigravity and Grok to DeepSeek Harness agents that appear as native subagents with live progress, tier/effort policy, escalation, and workspace guardrails. Also lends text-only harnesses vision and image generation through DeepSeek VL models or existing CLIs.354 npmMIT
- AlicenseNot gradedqualityAmaintenanceEnables coding hosts like Claude Code, Codex, Antigravity and Grok to dispatch work to DeepSeek Harness agents that appear as native subagents with live progress, tier/effort policy, escalation on failure, workspace guardrails and a jobs board. Also lends text-only harnesses vision and image generation via DeepSeek's VL model, the host's existing CLIs, or custom OpenAI-compatible endpoints.354 npmMIT
- AlicenseNot gradedqualityAmaintenanceEnables hosts like Claude Code, Codex, Antigravity and Grok to dispatch tasks to DeepSeek Harness agents while still showing them as native subagents with live progress, tier/effort presets, escalation, dispatch guardrails and a jobs board. Also lends text-only harnesses vision and image generation through DeepSeek's VL model or existing CLIs and custom OpenAI-compatible endpoints.354 npmMIT