How to config personal ai agent in vscode using Continue

Introduction

The Continue VS Code extension transforms your editor into an AI-powered coding companion by allowing you to define custom AI agents that assist with code generation, refactoring, documentation, and more. At the heart of this customization is the config.yaml file — a simple yet powerful configuration file that lets you tailor the behavior, model, prompts, and tools of your AI agent.

This guide walks you through creating a custom AI agent in VS Code using Continue’s config.yaml, enabling you to build a developer assistant tuned to your team’s coding standards, frameworks, or workflows.

Prerequisites

Visual Studio Code

Continue Extension installed from the VS Code Marketplace

An API key for a supported LLM provider (e.g., OpenAI, Anthropic, Ollama, or local models)

Step 1: Locate or Create `config.yaml`

When you first open Continue in VS Code, it automatically creates a .continue folder in your project root. Inside this folder, you’ll find a default config.yaml.

💡 If the folder doesn’t exist, open VS Code, press Ctrl+Shift+P (or Cmd+Shift+P), type Continue: Open Config, and select it. This will generate the folder and file.

The default config.yaml looks like this:

yaml


# .continue/config.yaml

models:
  - name: gpt-4-turbo
    provider: openai
    apiKey: ${{ secrets.MY_AI_SECRET_KEY }}

# Default agent settings
agent:
  maxTokens: 2048
  temperature: 0.3
  systemPrompt: |
    You are a helpful AI coding assistant. You write clean, efficient, and well-documented code.
    Always follow the project’s existing patterns and conventions.

# Custom commands (optional)
commands:
  - name: "Refactor for performance"
    description: "Refactor the selected code to improve performance and reduce complexity."
    prompt: |
      Analyze the following code and refactor it for better performance.
      Consider algorithmic efficiency, memory usage, and readability.
      Return only the improved code with no explanations.

Step 2: Secure Your API Key

Never hardcode API keys directly in config.yaml. Instead, store them securely in a .env file and load them as environment variables.

Create `.env`

Create a .env file in the same path of config.yaml (add it to .gitignore!):

json


  MY_AI_SECRET_KEY = xxxxxxxxxxxxxxxx

Add this line to your .gitignore to prevent leaks:
plain
.env
Plain text

Reload configuration

At top-right corner of Continue chat box, drop down Local Config, select Reload

Then you could use your private ai agent in vscode!

Continue will automatically read ${{MY_AI_SECRET_KEY}} from your environment — no secrets are stored in version control.

🔐 Security Note: Never commit secrets.json or .env files. Always use .gitignore to protect them.

Introduction

Prerequisites

Step 1: Locate or Create `config.yaml`

yaml

Step 2: Secure Your API Key

Create `.env`

json

plain

Reload configuration

Further Reading

JK Yang

Introduction

Prerequisites

Step 1: Locate or Create config.yaml

yaml

Step 2: Secure Your API Key

Create .env

json

plain

Reload configuration

Further Reading

JK Yang

Step 1: Locate or Create `config.yaml`

Create `.env`