2025-09-13 13:54:21 +08:00
2025-09-11 13:29:12 +00:00
2025-09-11 13:29:12 +00:00
2024-12-03 16:21:19 +08:00
2025-09-13 13:54:21 +08:00
2025-05-10 12:14:22 +08:00
2024-12-03 16:21:19 +08:00
2025-08-19 20:47:02 +08:00
2025-09-11 08:23:34 +00:00
2024-12-03 16:21:19 +08:00
2025-09-11 13:30:57 +00:00
2025-04-28 02:54:13 +00:00
2024-12-03 16:21:19 +08:00
2025-09-11 08:23:34 +00:00
2025-09-11 01:09:01 +08:00
2025-05-10 12:20:27 +08:00
2025-01-02 09:29:49 +08:00
2024-12-03 16:21:19 +08:00

Install

  1. 推荐使用Conda做虚拟环境管理自行百度在电脑上安装Conda(或miniconda)

  2. git clone https://gitee.com/CakeCN/code-agent.git

  3. conda create -n algoAgent python=3.10

  4. conda activate algoAgent

  5. 安装AgentScopecode-agent/AlgoriAgent> pip install -e .

  6. 安装其他依赖code-agent> pip install -r .\requirements.txt

  7. 最后启动服务器code-agent\Html> python app.py

  8. 然而还没结束为了让Vscode-web跑起来还需要安装code-server与code-server插件在公网场景下还要配置nginx反代(不过本地大家跑不需要配置nginx)

  9. 安装code-server [【Linux】浏览器写代码部署code-server远程vscode网页-CSDN博客](https://blog.csdn.net/muxuen/article/details/130334319)

  10. 我是在wsl上安装的启动code-server你应该能看到如下消息1746848009135

  11. 接下来配置好configVscode端口号要与实际端口一致

    1746850374313

  12. 大部分操作应该要登陆直接检查db文件夹下的内容即可cake123123

  13. Vscode服务器需要安装插件,Vscode-AgentHands-0.0.1.vsix通过Web安装vsix即可可以百度一下

启动

pip install "gunicorn>=20" eventlet

gunicorn -k eventlet -w 1 -b 0.0.0.0:5551 wsgi:app

config


[Global]
base_chat_url = https://www.dmxapi.com/v1
api_key = sk-aMpnWklN2IbsK44d1kNpy6YOP9bk1pdPJjFeEmbb0a5ytEFf
model=gpt-4.1-nano

[VSCODE_WEB]
url = https://hsamooc.cn
#http://asengine.net:8282
[VSCODE_WEB_PATH]
is_wsl = true
windows_path = F:/
wsl_path = /mnt/f/
[USER_DATA]
dir = db/data/user/
[COURSE_DATA]
dir = db/data/course/
[MONGO]
uri = mongodb://admin:12138cake@hsamooc.com:27017/hsamooc_test?authSource=admin

[TENCENT_COS]
secret_id = AKIDZ0EG4f2FE0YszsXzEF5h5GDwSlOtDLGx
secret_key = WqUv99vUpwqWzEF2kEVYNzNXTChy2wfB
bucket = hsamooc-cdn-1374354408
region = ap-guangzhou

[ASE_ENGINE]
url = https://asengine.net
namespace = /socket.io/c159a41d238b00f35ff33e036d007a0dec4d03197084b5557949a2f7f5ffce03


Description
No description provided
Readme AGPL-3.0 138 MiB
Languages
JavaScript 76.4%
HTML 10.9%
Python 9.3%
CSS 3.4%