CodeBuddy Code

CodeBuddy Code is a terminal AI programming tool (CLI) developed by Tencent Cloud CodeBuddy. It supports AI-powered conversations, code generation, file operations, and terminal command execution via the command line.

Prerequisites

• You have activated a TCB environment and obtained the environment ID cloudBaseEnvID
• Enable the required model in the AI console
• Obtain the Base URL and API Key in the AI console

Base URLhttps://<ENV_ID>.api.tcloudbasegateway.com/v1/ai/cloudbase

Configuration Steps

CodeBuddy Code manages custom models through a models.json configuration file.

Configuration field reference

Field	Description
id	Model ID, copy the enabled model identifier from the AI Console model list
name	Display name for the model
url	Model Base URL, obtained from the AI Console
apiKey	API Key for calling the model, obtained from the AI Console

User-level Config (Global)

1. Open or create the CodeBuddy Code configuration file

note

If browsing for the file, your file browser may hide files starting with ".". You may need to enable showing hidden files.

On Mac, use Finder shortcut Command + Shift + .. On Windows, check "File Explorer - View - Show - Hidden items".

• Mac/Linux path

~/.codebuddy/models.json

• Windows path

C:\Users\<your-username>\.codebuddy\models.json

2. Add the following configuration

{
  "models": [
    {
      "id": "hy3-preview",
      "name": "hy3-preview (CloudBase)",
      "vendor": "CloudBase",
      "url": "https://{{cloudBase-env-ID}}.api.tcloudbasegateway.com/v1/ai/cloudbase",
      "apiKey": "eyJhbGciOiJSUzI1N...",
      "maxInputTokens": 128000,
      "maxOutputTokens": 8192,
      "supportsToolCall": true,
      "supportsImages": true
    }
  ],
  "availableModels": ["hy3-preview"]
}

3. Start CodeBuddy Code

codebuddy --model hy3-preview

Or switch models during interaction with the /model command.

Project-level Config (Current Project)

Create .codebuddy/models.json in your project root. The format is the same as user-level config. Project-level config takes priority over user-level and overrides models with the same ID.

<project-root>/.codebuddy/models.json

The configuration content is the same as user-level config. CodeBuddy Code automatically reads project-level config on startup.

Troubleshooting

Error	Solution
Request failed / connection timeout	Verify the Base URL in the url field is correct
401 authentication failed	API Key is incorrect or expired, please obtain a new one from the AI Console
429 Token usage exceeded quota limit	Model name is incorrect, or the model is not enabled in the AI Console
Model not visible in list	Check that the availableModels array includes the model ID