Using Cline / RooCode
Cline (formerly Claude-dev) and RooCode are AI-powered coding assistants that help developers write, debug, and refactor code directly in their IDE. This guide shows how to configure these tools to use Azerion Intelligence as the underlying AI provider.
What are Cline and RooCode?
Cline is a VS Code extension that provides AI coding assistance for code generation, debugging, and refactoring. RooCode is a similar AI coding assistant with multi-language support for intelligent code suggestions and error detection.
Prerequisites
- VS Code installed (for Cline) or compatible IDE (for RooCode)
- Your Azerion Intelligence API key from https://app.azerion.ai/account#api-tokens
- The respective extension installed in your IDE
Integration Steps
Configuring Cline
-
Install the Extension:
- Open VS Code Extensions (Ctrl/Cmd + Shift + X)
- Search for "Cline" and install
-
Configure API Settings: Open VS Code Settings and add:
{
"cline.apiProvider": "openai-compatible",
"cline.apiUrl": "https://api.azerion.ai/v1",
"cline.apiKey": "YOUR_AZERION_API_KEY",
"cline.model": "meta.llama3-3-70b-instruct-v1:0"
}
Configuring RooCode
-
Configure API Connection: In your RooCode configuration file (
roocode.json
):{
"provider": "openai-compatible",
"baseUrl": "https://api.azerion.ai/v1",
"apiKey": "${ROOCODE_API_KEY}",
"model": "meta.llama3-3-70b-instruct-v1:0"
} -
Set Environment Variable:
# Linux/macOS
export ROOCODE_API_KEY="your_azerion_api_key"
# Windows PowerShell
$env:ROOCODE_API_KEY="your_azerion_api_key"
Basic Example
Once configured, open Cline chat in VS Code or activate RooCode in your IDE. Try this simple prompt:
Create a Python function that calculates the factorial of a number with error handling for negative inputs.
The AI will generate code using Azerion Intelligence's meta.llama3-3-70b-instruct-v1:0
model and provide explanations for the implementation.
Troubleshooting
Connection Issues
- Problem: "API connection failed"
- Solutions:
- Verify your API key is valid and active
- Check the base URL:
https://api.azerion.ai/v1
- Test connectivity with curl:
curl -H "Authorization: Bearer YOUR_API_KEY" https://api.azerion.ai/v1/models
Model Availability
- Problem: "Model not found" error
- Solutions:
- Ensure model name is exactly:
meta.llama3-3-70b-instruct-v1:0
- Use the List Models endpoint to verify availability
- Check your API key has access to the model
- Ensure model name is exactly:
Rate Limiting
- Problem: "Rate limit exceeded"
- Solutions:
- Reduce request frequency in your development workflow
- Consider upgrading your Azerion Intelligence plan
- Use more specific prompts to reduce token usage