← Back to Skills Marketplace
yang1002378395-cmyk

Deepseek Quick Config

by yang1002378395-cmyk · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
219
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install deepseek-quick-config
Description
Quickly configure DeepSeek V3 as the AI model for OpenClaw with API key setup, cost-effective Chinese optimization, and optional streaming or temperature tun...
README (SKILL.md)

DeepSeek Quick Config for OpenClaw

5 分钟配置 DeepSeek V3 作为 OpenClaw 的 AI 模型。

为什么选 DeepSeek?

  • 性价比最高:¥0.27/百万 input tokens vs GPT-4o ¥35/百万
  • 性能强劲:接近 GPT-4 水平
  • 中文优化:国产模型,中文理解更好
  • 无需 VPN:国内直连,速度快

快速配置

Step 1: 获取 API Key

  1. 访问 https://platform.deepseek.com/
  2. 注册/登录
  3. 进入 API Keys 页面
  4. 创建新 Key

Step 2: 配置 OpenClaw

# 方法 1:命令行配置
openclaw configure --section models

# 选择 DeepSeek,粘贴 API Key

或编辑 ~/.openclaw/config.yaml:

models:
  default: deepseek-chat
  providers:
    deepseek:
      apiKey: sk-xxxxxxxx
      baseURL: https://api.deepseek.com/v1

Step 3: 测试连接

openclaw chat "你好,介绍一下你自己"

成本对比

模型 Input (¥/百万) Output (¥/百万) 相对成本
DeepSeek V3 ¥0.27 ¥1.08 1x
GPT-4o mini ¥1.05 ¥4.20 4x
Claude Haiku ¥0.63 ¥3.15 3x
GPT-4o ¥35.00 ¥105.00 130x
Claude Sonnet ¥21.00 ¥105.00 97x

常见问题

Q: 提示 "API key invalid"

A: 检查 Key 是否正确复制,没有多余空格

Q: 响应很慢

A: 深度思考模式 (deepseek-reasoner) 会慢一些,日常用 deepseek-chat

Q: 想用 R1 深度思考?

A: 把模型名改成 deepseek-reasoner

高级配置

启用流式输出

models:
  providers:
    deepseek:
      stream: true

自定义温度

models:
  providers:
    deepseek:
      temperature: 0.7  # 0-2,默认 1

需要帮助?

  • 微信:yanghu_ai
  • Telegram: @yanghu_openclaw

Version: 1.0.0 Created: 2026-03-21

Usage Guidance
This is a simple how-to for wiring a DeepSeek API key into your OpenClaw config. Before installing/using it: (1) verify you trust platform.deepseek.com and understand their pricing (the guide highlights cost differences); (2) be aware the guide tells you to place your API key in ~/.openclaw/config.yaml (plaintext) — consider using a secrets manager or environment variables if you prefer not to store keys on disk; (3) granting the agent access to use the configured model means the agent can make API calls that consume the key and incur charges, so review agent permissions and usage limits; (4) the SKILL.md includes personal contact handles — that is external and unrelated to functionality, so treat them as optional support channels. Overall the skill is coherent and low-risk, but protect your API key and monitor usage after enabling the model.
Capability Analysis
Type: OpenClaw Skill Name: deepseek-quick-config Version: 1.0.0 The bundle is purely informational, containing documentation (SKILL.md) and metadata (_meta.json, package.json) for manually configuring the DeepSeek AI model in OpenClaw. It contains no executable code, automated scripts, or malicious instructions.
Capability Assessment
Purpose & Capability
The name/description match the content: SKILL.md shows steps to obtain a DeepSeek API key and add it to OpenClaw's config or use the openclaw configure command. There are no unrelated environment variables, binaries, or install steps.
Instruction Scope
Instructions are narrowly scoped to visiting DeepSeek, creating an API key, editing ~/.openclaw/config.yaml (or using the CLI), and testing the model. They reference only the DeepSeek endpoints (platform.deepseek.com, api.deepseek.com) and the OpenClaw config, which is appropriate for the stated task. Note: the guide directs storing an API key in a plaintext config file, which is expected but has the usual secret-storage risks.
Install Mechanism
No install spec or code is provided (instruction-only). Nothing is downloaded or written to disk by the skill itself, minimizing install-time risk.
Credentials
The skill declares no environment variables and asks the user to provide only a DeepSeek API key in the OpenClaw config — proportional to the goal. As noted, this stores a secret locally; consider using protected storage or environment variables if preferred.
Persistence & Privilege
always is false and the skill is instruction-only; it does not request permanent presence, modify other skills, or access unrelated credentials. Be aware that if you configure OpenClaw to use DeepSeek, the agent may send requests to DeepSeek using that API key and incur costs.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install deepseek-quick-config
  3. After installation, invoke the skill by name or use /deepseek-quick-config
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
DeepSeek Quick Config for OpenClaw v1.0.0 - Added concise setup instructions to use DeepSeek V3 as the AI model with OpenClaw in under 5 minutes. - Included a step-by-step guide for obtaining an API key and configuring via command line or config file. - Provided detailed pricing comparison charts versus leading AI models. - Added troubleshooting FAQ and tips for advanced configuration like streaming and temperature settings. - Listed support contact information for WeChat and Telegram.
Metadata
Slug deepseek-quick-config
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Deepseek Quick Config?

Quickly configure DeepSeek V3 as the AI model for OpenClaw with API key setup, cost-effective Chinese optimization, and optional streaming or temperature tun... It is an AI Agent Skill for Claude Code / OpenClaw, with 219 downloads so far.

How do I install Deepseek Quick Config?

Run "/install deepseek-quick-config" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Deepseek Quick Config free?

Yes, Deepseek Quick Config is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Deepseek Quick Config support?

Deepseek Quick Config is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Deepseek Quick Config?

It is built and maintained by yang1002378395-cmyk (@yang1002378395-cmyk); the current version is v1.0.0.

💬 Comments