ClaudeCode+KIMI+CC-Connect打造轻量级Openclaw

Openclaw最近的更新有点爆炸，bug太多，升级很不友好

这里推荐一个全新的解决方案：ClaudeCode+KIMI+CC-Connect 完美取代Openclaw

以下针对Windows环境，此外要会魔法上网用于下载Claude Code命令行。

下载安装Claude Code命令行

Windows PowerShell:

复制代码

irm https://claude.ai/install.ps1 | iex

Windows CMD:

复制代码

curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd

claude会默认安装到 "C:\Users\你的用户名\.local\bin" 这个路径下，确保把这个路径加到环境变量里。

然后启动命令行测试：

下载安装CC-Connect

CC-Connect的作用相当于一个桥接器，让IM软件比如钉钉、飞书能和你的claudecode沟通。

同时他会根据配置，让claudecode调用你设置的大模型，比如Minimax或KIMI，而不是Anthropic的LLM。这样之后使用的时候，就不用魔法上网了。

具体做法很简单：

去https://github.com/chenhg5/cc-connect/releases 下载amd64的压缩包，打开压缩包，运行里面的exe文件。

成功后，一个网站会运行在http://127.0.0.1:9820/

目前有些bug，默认只用feishu作为IM，如果要改动去 "C:\Users\用户名\.cc-connect" 修改config.toml文件。

可以在服务商里添加国内的大模型，KIMI 和 MINIMAX都用兼容 anthropic claude code的API接口。

然后记得在项目中启用服务商：

最后给到一段参考的config.toml:

复制代码

data_dir = ""
attachment_send = ""
language = "en"

[[providers]]
  name = "claudecode-kimi"
  api_key = "sk-xxxxxxxxxxxxxxx"
  base_url = "https://api.moonshot.cn/anthropic"
  model = "kimi-k2.5"
  agent_types = ["claudecode"]

  [[providers.models]]
    model = "kimi-k2.5"

  [[providers.models]]
    model = "kimi-k2.6"

[[projects]]
  name = "my-project"

  [projects.agent]
    type = "claudecode"
    provider_refs = ["claudecode-kimi"]

    [projects.agent.options]
      mode = "default"
      work_dir = "C:\\Users\\你的用户名\\工作目录"

provider = "claudecode-kimi"

  [[projects.platforms]]
    type = "dingtalk"

    [projects.platforms.options]
      client_id = "dingxxxx"
      client_secret = "xxxxxx"

[log]
  level = "info"

[speech]
  enabled = false
  provider = ""
  language = ""

  [speech.openai]
    api_key = ""
    base_url = ""
    model = ""

  [speech.groq]
    api_key = ""
    model = ""

  [speech.qwen]
    api_key = ""
    base_url = ""
    model = ""

  [speech.gemini]
    api_key = ""
    model = ""

[tts]
  enabled = false
  provider = ""
  voice = ""
  tts_mode = ""
  max_text_len = 0

  [tts.openai]
    api_key = ""
    base_url = ""
    model = ""

  [tts.qwen]
    api_key = ""
    base_url = ""
    model = ""

  [tts.minimax]
    api_key = ""
    base_url = ""
    model = ""

[cron]
  session_mode = ""

[webhook]
  port = 0

[bridge]
  enabled = true
  port = 9810
  token = "xxx"
  cors_origins = ["*"]

[management]
  enabled = true
  port = 9820
  token = "xxxx"
  cors_origins = ["*"]