diff --git a/README.md b/README.md index c349dcd..af516b0 100644 --- a/README.md +++ b/README.md @@ -191,9 +191,15 @@ English/[简体中文](https://github.com/deepseek-ai/awesome-deepseek-integrati + + + + + + diff --git a/README_cn.md b/README_cn.md index 56fbe04..a500a1d 100644 --- a/README_cn.md +++ b/README_cn.md @@ -146,6 +146,11 @@ ### AI Agent 框架
Icon DeepSeek MCP Server Model Context Protocol server for DeepSeek's advanced language models.
Icon SuperAgentX SuperAgentX: A Lightweight Open Source AI Framework Built for Autonomous Multi-Agent Applications with Artificial General Intelligence (AGI) Capabilities.
Icon
+ + + + diff --git a/docs/model_context_protocol/README.md b/docs/model_context_protocol/README.md new file mode 100644 index 0000000..7b3b74d --- /dev/null +++ b/docs/model_context_protocol/README.md @@ -0,0 +1,131 @@ + +# DeepSeek MCP Server + +A Model Context Protocol (MCP) server for the DeepSeek API, allowing seamless integration of DeepSeek's powerful language models with MCP-compatible applications like Claude Desktop. + +## *Anonymously* use DeepSeek API -- Only a proxy is seen on the other side + +DeepSeek Server MCP server +Smithery Badge + + +[![npm version](https://img.shields.io/npm/v/deepseek-mcp-server)](https://www.npmjs.com/package/deepseek-mcp-server) +[![npm downloads](https://img.shields.io/npm/dm/deepseek-mcp-server)](https://www.npmjs.com/package/deepseek-mcp-server) +[![GitHub issues](https://img.shields.io/github/issues/DMontgomery40/deepseek-mcp-server)](https://github.com/DMontgomery40/deepseek-mcp-server/issues) +[![GitHub forks](https://img.shields.io/github/forks/DMontgomery40/deepseek-mcp-server)](https://github.com/DMontgomery40/deepseek-mcp-server/network) +[![GitHub stars](https://img.shields.io/github/stars/DMontgomery40/deepseek-mcp-server)](https://github.com/DMontgomery40/deepseek-mcp-server/stargazers) +[![GitHub license](https://img.shields.io/github/license/DMontgomery40/deepseek-mcp-server?color=blue)](https://github.com/DMontgomery40/deepseek-mcp-server/blob/main/LICENSE) + +## Installation + +### Installing via Smithery + +To install DeepSeek MCP Server for Claude Desktop automatically via [Smithery](https://smithery.ai/server/@dmontgomery40/deepseek-mcp-server): + +```bash +npx -y @smithery/cli install @dmontgomery40/deepseek-mcp-server --client claude +``` + +### Manual Installation +```bash +npm install -g deepseek-mcp-server +``` +### Usage with Claude Desktop + +Add this to your `claude_desktop_config.json`: + +```json +{ + "mcpServers": { + "deepseek": { + "command": "npx", + "args": [ + "-y", + "deepseek-mcp-server" + ], + "env": { + "DEEPSEEK_API_KEY": "your-api-key" + } + } + } +} +``` + +## Features + +> Note: The server intelligently handles these natural language requests by mapping them to appropriate configuration changes. You can also query the current settings and available models: + +- User: "What models are available?" + - Response: Shows list of available models and their capabilities via the models resource. +- User: "What configuration options do I have?" + - Response: Lists all available configuration options via the model-config resource. +- User: "What is the current temperature setting?" + - Response: Displays the current temperature setting. +- User: "Start a multi-turn conversation. With the following settings: model: 'deepseek-chat', make it not too creative, and + allow 8000 tokens." + - Response: *Starts a multi-turn conversation with the specified settings.* + +### Automatic Model Fallback if R1 is down + +- If the primary model (R1) is down (called `deepseek-reasoner` in the server), the server will automatically attempt to try with v3 (called `deepseek-chat` in the server) +> Note: You can switch back and forth anytime as well, by just giving your prompt and saying "use `deepseek-reasoner`" or "use `deepseek-chat`" +- V3 is recommended for general purpose use, while R1 is recommended for more technical and complex queries, primarily due to speed and token useage + +### Resource discovery for available models and configurations: + * Custom model selection + * Temperature control (0.0 - 2.0) + * Max tokens limit + * Top P sampling (0.0 - 1.0) + * Presence penalty (-2.0 - 2.0) + * Frequency penalty (-2.0 - 2.0) + +## Enhanced Conversation Features + +**Multi-turn conversation support:** +* Maintains complete message history and context across exchanges +* Preserves configuration settings throughout the conversation +* Handles complex dialogue flows and follow-up chains automatically + +This feature is particularly valuable for two key use cases: + +1. **Training & Fine-tuning:** + Since DeepSeek is open source, many users are training their own versions. The multi-turn support provides properly formatted conversation data that's essential for training high-quality dialogue models. + +2. **Complex Interactions:** + For production use, this helps manage longer conversations where context is crucial: + * Multi-step reasoning problems + * Interactive troubleshooting sessions + * Detailed technical discussions + * Any scenario where context from earlier messages impacts later responses + +The implementation handles all context management and message formatting behind the scenes, letting you focus on the actual interaction rather than the technical details of maintaining conversation state. + + + + +## Testing with MCP Inspector + +You can test the server locally using the MCP Inspector tool: + +1. Build the server: + ```bash + npm run build + ``` + +2. Run the server with MCP Inspector: + ```bash + # Make sure to specify the full path to the built server + npx @modelcontextprotocol/inspector node ./build/index.js + ``` + +The inspector will open in your browser and connect to the server via stdio transport. You can: +- View available tools +- Test chat completions with different parameters +- Debug server responses +- Monitor server performance + +Note: The server uses DeepSeek's R1 model (deepseek-reasoner) by default, which provides state-of-the-art performance for reasoning and general tasks. + +## License + +MIT diff --git a/docs/model_context_protocol/README_cn.md b/docs/model_context_protocol/README_cn.md new file mode 100644 index 0000000..0c792c6 --- /dev/null +++ b/docs/model_context_protocol/README_cn.md @@ -0,0 +1,128 @@ +# DeepSeek MCP 服务器 + +这是一个适用于 DeepSeek API 的 Model Context Protocol (MCP) 服务器，可与 Claude Desktop 等兼容 MCP 的应用程序无缝集成，从而利用 DeepSeek 强大的语言模型。 + +## *匿名* 使用 DeepSeek API —— 另一端只会看到代理 + +DeepSeek MCP Server +Smithery Badge + +[![npm version](https://img.shields.io/npm/v/deepseek-mcp-server)](https://www.npmjs.com/package/deepseek-mcp-server) +[![npm downloads](https://img.shields.io/npm/dm/deepseek-mcp-server)](https://www.npmjs.com/package/deepseek-mcp-server) +[![GitHub issues](https://img.shields.io/github/issues/DMontgomery40/deepseek-mcp-server)](https://github.com/DMontgomery40/deepseek-mcp-server/issues) +[![GitHub forks](https://img.shields.io/github/forks/DMontgomery40/deepseek-mcp-server)](https://github.com/DMontgomery40/deepseek-mcp-server/network) +[![GitHub stars](https://img.shields.io/github/stars/DMontgomery40/deepseek-mcp-server)](https://github.com/DMontgomery40/deepseek-mcp-server/stargazers) +[![GitHub license](https://img.shields.io/github/license/DMontgomery40/deepseek-mcp-server?color=blue)](https://github.com/DMontgomery40/deepseek-mcp-server/blob/main/LICENSE) + +## 安装 + +### 通过 Smithery 安装 + +要使用 Smithery 在 Claude Desktop 上自动安装 DeepSeek MCP Server，请执行以下命令（请确保已安装 `@smithery/cli`）： + +`bash +npx -y @smithery/cli install @dmontgomery40/deepseek-mcp-server --client claude +` + +### 手动安装 + +`bash +npm install -g deepseek-mcp-server +` + +### 在 Claude Desktop 中使用 + +在你的 `claude_desktop_config.json` 中添加： + +`json +{ + "mcpServers": { + "deepseek": { + "command": "npx", + "args": [ + "-y", + "deepseek-mcp-server" + ], + "env": { + "DEEPSEEK_API_KEY": "your-api-key" + } + } + } +} +` + +## 功能简介 + +> 注意：该服务器能够根据自然语言请求智能地将其映射到相应的配置更改。你也可以查询当前设置和可用模型： + +- 用户：“有哪些可用的模型？” + - 响应：通过 models 资源列出可用模型及其功能。 +- 用户：“我有哪些配置选项？” + - 响应：通过 model-config 资源列出所有可用的配置选项。 +- 用户：“当前的温度（temperature）设置是多少？” + - 响应：显示当前温度设置。 +- 用户：“开始一个多轮对话。使用如下设置：model: 'deepseek-chat'，创意度不要太高，并且允许 8000 个 token。” + - 响应：使用指定设置启动一个多轮对话。 + +### 当 R1 出现故障时自动回退到其他模型 + +- 如果主模型（R1，服务器中称为 `deepseek-reasoner`）出现故障，服务器会自动尝试使用 v3（服务器中称为 `deepseek-chat`） +- 你也可以随时在对话中切换，只需在对话中输入提示并说“使用 `deepseek-reasoner`”或“使用 `deepseek-chat`” +- v3 更适用于通用场景；R1 更适用于处理较为复杂的技术性问题，主要得益于速度和 token 使用的优化 + +### 资源发现：可用的模型和配置 + +- 自定义模型选择 +- 温度控制（0.0 - 2.0） +- 最大 token 限制 +- Top P 采样（0.0 - 1.0） +- 存在惩罚（presence penalty）（-2.0 - 2.0） +- 频率惩罚（frequency penalty）（-2.0 - 2.0） + +## 增强的对话功能 + +**多轮对话支持：** +- 在多轮交互过程中维护完整的消息历史和上下文 +- 在对话过程中保留配置设置 +- 自动处理复杂的对话逻辑和后续请求 + +这一功能在以下两个主要场景中特别有价值： + +1. **训练 & 微调：** + - 由于 DeepSeek 是开源的，很多用户正在训练自己的版本。多轮对话支持能够提供格式正确的对话数据，这对于训练高质量对话模型至关重要。 + +2. **复杂场景交互：** + - 在生产环境中，这种功能有助于管理需要保留上下文的更长对话，例如： + * 多步骤推理问题 + * 交互式故障排查 + * 详尽的技术讨论 + * 任何需要利用早期消息上下文来影响后续响应的场景 + +该功能在幕后自动处理所有上下文管理和消息格式，你只需关注对话本身，无需担心维护对话状态的技术细节。 + +## 使用 MCP Inspector 进行测试 + +你可以使用 MCP Inspector 工具在本地测试服务器： + +1. 构建服务器： + `bash + npm run build + ` + +2. 使用 MCP Inspector 启动服务器： + `bash + npx @modelcontextprotocol/inspector node ./build/index.js + ` + +MCP Inspector 将在你的浏览器中打开，并通过 stdio 传输连接到该服务器。你可以： + +- 查看可用工具 +- 使用不同参数测试对话补全 +- 调试服务器响应 +- 监控服务器性能 + +注意：服务器默认使用 DeepSeek 的 R1 模型（`deepseek-reasoner`），它在推理和通用任务方面具有最先进的性能。 + +## 许可证 + +MIT
Icon DeepSeek MCP Server 用于 DeepSeek 高级语言模型的 Model Context Protocol 服务器 +
Icon Anda